Nerve impulses can collide and continue unaffected

According to the traditional theory of nerves, two nerve impulses sent from opposite ends of a nerve annihilate when they collide. New research from the Niels Bohr Institute now shows that two colliding nerve impulses simply pass through each other and continue unaffected. This supports the theory that nerves function as sound pulses. The results are published in the scientific journal Physical Review X.

Nerve signals control the communication between the billions of cells in an organism and enable them to work together in neural networks. But how do nerve signals work?

Old model

In 1952, Hodgkin and Huxley introduced a model in which nerve signals were described as an electric current along the nerve produced by the flow of ions. The mechanism is produced by layers of electrically charged particles (ions of sodium and potassium) on either side of the nerve membrane that change places when stimulated. This change in charge creates an electric current.

This model has enjoyed general acceptance. For more than 60 years, all medical and biology textbooks have said that nerves function is due to an electric current along the nerve pathway. However, this model cannot explain a number of phenomena that are known about nerve function.

New model

Researchers at the Niels Bohr Institute at the University of Copenhagen have now conducted experiments that raise doubts about this well-established model of electrical impulses along the nerve pathway.

“According to the theory of this ion mechanism, the electrical signal leaves an inactive region in its wake, and the nerve can only support new signals after a short recovery period of inactivity. Therefore, two electrical impulses sent from opposite ends of the nerve should be stopped after colliding and running into these inactive regions,” explains Thomas Heimburg, Professor and head of the Membrane Biophysics Group at the Niels Bohr Institute at the University of Copenhagen.

Thomas Heimburg and his research group conducted experiment in the laboratory using nerves from earthworms and lobsters. The nerves were removed and used in an experiment which allowed the researchers to stimulate the nerve fibres with electrodes on both ends. Then they measured the signals en route. 

“Our study showed that the signals passed through each other completely unhindered and unaltered. That’s how sound waves work. A sound wave doesn’t stop when it meets another sound wave. Both waves continue on unimpeded. The nerve impulse can therefore be explained by the fact that the pulse is a mechanical wave in the form of a sound pulse, a soliton, that moves along the nerve membrane,” explains Thomas Heimburg.

The theory is confirmed

When the sound pulse moves through the nerve pathway, the membrane changes locally from a liquid to a more solid form. The membrane is compressed slightly, and this change leads to an electrical pulse as a consequence of the piezoelectric effect.

“The electrical signal is thus not based on an electric current but is caused by a mechanical force,” points out Thomas Heimburg.

Thomas Heimburg, along with Professor Andrew Jackson, first proposed the theory that nerves function by sound pulses in 2005. Their research has since provided support for this theory, and the new experiments offer additional confirmation for the theory that nerve signals are sound pulses.

(Image caption: A thalamocortical, or TC neuron labeled with fluorescent dye, as used in Dr. Augustinaite’s study. The image shows a voltage recording device, at bottom left, entering the yellow cell body, and a stimulation device, at top, reaching the dendrites. Color in this image shows the depth in the slice.)

To See or Not to See

The brain is a complicated network of small units called neurons, all working to carry information from the outside world, create an internal model, and generate a response. Neurons sense a signal through branching dendrites, carry this signal to the cell body, and send it onwards through a long axon to signal the next neuron. However, neurons can function in many different ways; some of which researchers are still exploring. Some signals that the dendrites receive do not continue to the next neuron; instead they seem to change the way that the neuron handles the subsequent signals. This could help neurons function as part of a large network, but researchers still have many questions. Dr. Sigita Augustinaite, a researcher in the Optical Neuroimaging Unit at the Okinawa Institute of Science and Technology Graduate University, suggested one mechanism explaining how neurons help the network function. Her findings, part of collaboration between the University of Oslo and OIST, were published August 13, 2014 as the cover article in The Journal of Neuroscience.

Dr. Augustinaite studies the visual pathway, where signals from the retina are sent to the visual cortex, where the brain interprets signals from the eye. Between the eye and the visual cortex, the signals must pass through the visual thalamus, that is, through thalamocortical, or TC neurons. These neurons can switch between a “sleeping” state and a “waking” state depending on input they receive from neurons and other brain areas. When an animal is awake, TC neurons transmit the incoming retinal signals on to the cortex, but when the animal is asleep, the neurons block retinal signals.

The visual cortex also sends a massive input back to TC neurons to control retinal signals traveling through the thalamus. But Dr. Augustinaite says that the suggested mechanisms of this control bring more questions than answers. To understand more, she conducted experiments in acute brain slices, small pieces of brain tissue where neurons stay alive and maintain their physiological properties. She added glutamate to dendrites far from the cell body to emulate a feedback signal from the visual cortex. Then she measured the neuron’s response, shown as a voltage difference between inside and outside of the membrane.

Dr. Augustinaite found that stimulating the neurons in this way depolarizes their membranes, creating something called NMDA spike/plateau potentials. If strong enough, depolarization can cause a neuron to fire an action potential, which travels through the axon to activate other neurons. Action potentials look like a sharp, one-millisecond increase in membrane voltage, and they transmit signals from retina to cortex. But if NMDA spike/plateaus induces action potentials, signals from the cortex and signals from the retina would be indistinguishable. With her experiments, Dr. Augustinaite showed that the NMDA spike/plateau potentials in TC neurons do not trigger action potentials. Instead, they lift the voltage of the membrane, changing the neuron’s properties for few hundred milliseconds, creating conditions for reliable signal transmission from retina to cortex.

“The research gives, for the first time, a clear view on what dendritic potentials are good for,” explained Prof. Bernd Kuhn, who leads the lab where Dr. Augustinaite works. “It points directly to the mechanism,” he concluded. Showing how dendritic plateaus function is just one important step toward understanding how neurons function as a network. “This mechanism could also be used in many other neuronal circuits, where one input regulates how another input moves through the network,” Dr. Augustinaite said. “This mechanism is an exciting logical element in the neuronal network, but just the start of putting the puzzle together.”

Pulse propagation and signal transduction in the hydraulic brain

When Descartes turned his critical eye to the nervous system, he reasoned that the nerves must transduce hydraulic power to control the musculature. In the circulatory system, blood is pushed comparatively slowly through the aorta, perhaps around 0.3 meters per second. Superimposed on that flow, however, is an arterial pulse wave which propagates much faster, both through the blood and the walls of the vessel. For compliant and healthy vessels that speed might be around 10 meters per second, while for more hardened arteries, it could be 15 or higher. Modern day electrophysiologists have since replaced the hydraulic model with the idea that nerves really only transmit information—electrical information no less. Yet when looking at the power supply to the leg, for example, it is still hard to ignore the fact that the main femoral artery, at a diameter scarcely a half of an inch, looks rather meager next to the “information-supplying” sciatic nerve, which may actually be more like three-quarters of an inch. A conflux of ideas from a variety of disciplines has recently led to a critical re-emergence of the more mechanical side of the nervous system. To that point, two German scientists have just published a paper in the journal, Medical Hypotheses, where they suggest that the pulse wave is the main event in nervous conduction, while the electrical show is mere epiphenomenon.

We recently discussed the increasingly popular idea that action potentials may actually be soliton waves which propagate in the membranes of axons as phase transitions with minimal loss in energy. Convincing biologists that these subtle creatures could exist in the chaotic and varied conditions inside neurons has been a challenge. However, it is harder to argue against the fact that any kind of electrochemical spike based on the rapid influx of ions will be accompanied by a significant pressure pulse. The idea that the German researchers have supported, is that these as the pressure pulses naturally decay in the viscoelastic medium of the nerve, they are refreshed by ionic input at the nodes between myelinated axon segments, or continuously in unmyelinated axons.

If you have ever been absent-minded enough to grab a live wire, or even brush up strongly against one, the sensation is unforgettable. It is not such a stretch to acknowledge that when you slam your funny bone, or more precisely the Ulnar nerve (largest unprotected nerve in your body), the resultant vibe and decay feels almost identical to a real electrical assault. Similarly, the so-called “stingers” that run down the limbs after a sharp blow to the head are familiar to most footballers, and can give one quite a shock. Unfortunately these (albeit very simplistic) macroscopic intuitions don’t hold up so well when extended to the microscopic domain. Granted, when the electrochemical mechanisms that are assumed to underlie nervous conduction are looked at in detail, it becomes more difficult to disentangle the mechanical from the electric. However, as the authors observe, at some point, an attentive electrophysiologist must ask his or herself, “why are so many ion channels mechanosensitive” ?

One unexpected finding of the patch clamp recording technique was that the dilation of the membrane caused by local tension leads to considerable increase in transmembrane ion flow. Impulse waves causing short extensions in the membrane can directly induce opening and closing of both voltage and ligand gated channels. The idea that the pore in these channels is a rigid tube isolated from larger membrane events is difficult to support in this context. According to the authors, it is quite likely that common mechanoreceptor devices, like the pressure- or vibration-sensitive Vater-Pacinian corpuscles of the skin, conduct signals to initiate high-speed polysynaptic muscle reflex circuits without any classical intermediary electrical conversion.

The exact conduction velocity of mechanical impulses in nerve fibers remains unknown. It is estimated that under physiologic conditions, an unamplified axoplasmic pressure pulse would decay over roughly 1 mm due to viscosity, depending on the distensibility of the axon wall. When compared to the theoretical case of an absolutely rigid wall, a typical myelin sheath may be rigid enough to support pulse speeds up to one-fifth of the estimated maximum. That speed is not to shabby when compared with some rough estimates from previous authors, which put the maximum pulse velocity under an indistensible membrane somewhere upwards of 1500 meters per second. Suddenly, the quicker than life eyeblink response, or speed of the tenderfoot stepping on a sharp shard, become a little more comprehensible.

The theory as it stands is incomplete and needs to be adapted for specific cases with real biology in mind. In different animals, and in different regions of their brains, conduction in neurons goes by different names. For example, in the cerebellum, the unmyelinated parallel fibers pack to extreme densities in a regular crystalline lattice whose reason to be defies physiologic explanation to this day. Just as we currently have no good explanation for how signals could be properly isolated in nerve bundles where seemingly random nodes of Ranvier overlap in extent and influence, it is hard to imagine parallel fibers could maintain their electrochemical, or even mechanical, autonomy within this geometry.

The pressure wave theory wields considerable predictive power when it comes to explaining some of the unique synaptic specializations found throughout the brain. When considered only from an electrochemical point of view, the huge structural synaptic investments, like those found at the neuromuscular junction (NMJ), can hardly be imagined to be driven by local, and weak, current or field effects. One might need look no further than simple-to-recreate Chaldni patterns set up in two dimensions on the surface of a taunt drum, to make the imaginative leap to a three dimensional system, with multiple vibrating players, where more extreme patterns might easily be set up to provide authorship to repeatable complex structure. For the NMJ in particular, the case has been made that at the end-plate, the comparatively enormous efflux of acetylcholine to the deeply-guttered cleft, and propagation of excitation through the transverse tubule system, are all components of a continuous mechanical amplifier.

The apparent ease with which evolving organisms manage to cobble together all manner of sensitive hearing devices becomes infinitely more explicable once we see that nature has apparently been doing this kind of thing all alone inside of neurons. The amplification and transduction through liquid channels, of barely noticeable vibrations against a background of thermal noise much greater in magnitude, is in this light, no evolutionary stumble-upon, but rather the bread and butter of neural systems, and perhaps many aspects of life in general.

Wasp has hints of a clockwork brain

The greenhouse whitefly parasite (Encarsia formosa) is just half a millimetre in length. It parasitises the larvae of whiteflies and so it has long been used as a natural pest-controller.

To find out how its neurons have adapted to miniaturisation, Reinhold Hustert of the University of Göttingen in Germany examined the insect’s brain with an electron microscope. The axons - fibres that shuttle messages between neurons - were incredibly thin. Of 528 axons measured, a third were less than 0.1 micrometre in diameter, an order of magnitude narrower than human axons. The smallest were just 0.045 μm (Arthropod Structure & Development, doi.org/jfn).

That’s a surprise, because according to calculations by Simon Laughlin of the University of Cambridge and colleagues, axons thinner than 0.1 μm simply shouldn’t work. Axons carry messages in waves of electrical activity called action potentials, which are generated when a chemical signal causes a large number of channels in a cell’s outer membrane to open and allow positively charged ions into the axon. At any given moment some of those channels may open spontaneously, but the number involved isn’t enough to accidentally trigger an action potential, says Laughlin - unless the axon is very thin. An axon thinner than 0.1 μm will generate an action potential if just one channel opens spontaneously (Current Biology, doi.org/frfwpz).

"That makes the axon impossibly noisy," Laughlin says. Any "legitimate" action potentials will be drowned out.

Hustert suggests that a neuron might get around this problem by firing bursts of action potentials to cut through the noise, but Laughlin is sceptical. “They’d be firing furiously all the time,” he says, and every action potential costs energy.

Instead, the neurons might not bother with conventional action potentials at all. “They could be sending signals mechanically,” Laughlin says. The tiny axons might each carry a long rigid rod stretching down the centre. Pulling the rod could create a physical rather than electrical trigger for the release of a chemical that passes the signal on to the neighbouring neuron.

In larger animals this would be far too slow, says Laughlin, but in the tiny body of the greenhouse whitefly parasite, a partly “clockwork” brain might be the best approach.