I'm reading a book named Neuroscience:Exploring the Brain recently. And in the chapter about synaptic transmission, it says "Transmission at electrical synapses is very fast and, if the synapse is large, fail-safe." So what does fail-safe mean? And why is that transmission fail-safe when the synapse is large? And when is comes to EPSP, it also mentions that "The neuromuscular junction has evolved to be fail-safe; it needs to work every time, and the best way to ensure this is to generate an EPSP of a huge size." I don't understand the meaning of the word and the whole sentence.


1 Answer 1


To understand what failure means in this context, we need to know a bit about how signals are communicated from one neuron to another.

Types of signals

First of all, it's important to know that synaptic signaling (communication between neurons) comes in two flavors: chemical and electrical. The signaling is done at the interface between two neurons, known as the synapse. There are generally dozens to hundreds of synapses between two connected neurons. At chemical synapses, a chemical (the neurotransmitter) is released into the space between the pre-synaptic site on the axon and the post-synaptic site on the dendrite (or soma). At electrical synapses, no chemicals are released. Instead, porous proteins called “gap junction channels” bridge the space between the cell membranes of two neurons and create pores that allow electrical current to pass directly from one neuron to the other.

Signaling is probabilistic

Second, it's important to appreciate that neurotransmission is probabilistic, rather than deterministic. That just means there’s a chance that signaling from an upstream neuron might fail to evoke a response in a downstream neuron. There are several reasons why transmission is not 100% efficient, at both the pre-synaptic and post-synaptic ends.

Some of these reasons have to do with spontaneous opening and closing of channels in the membrane. When researchers examine electrical currents flowing through single ion channel proteins they find that each channel randomly flips between open and closed states. This is caused by molecular-scale fluctuations in the protein's shape that cause it to either block or allow the passage of electrical current (or ions).

In addition, random fluctuations in how well electrical current is able to be integrated by the downstream neuron add to the variability in signaling that makes responses probabilistic. You probably already know that "response" of a neuron refers to the firing of action potentials. And you probably also know that action potentials are generated when the membrane of a neuron becomes sufficiently depolarized - the result of a process known as synaptic integration. A simple view of synaptic integration is the summing together of all electrical currents flowing across the membrane from different synapses in a short window of time. If this current integrates to a total greater than some value (the firing threshold) the neuron fires, otherwise it does not. What some may not realize, though, is that this process isn't perfect - current can leak out of the cell, other neurons can "siphon off" current, and receptor proteins in the downstream neuron can become inactive and stop contributing to the overall current flux.

All of these factors conspire together to make the signaling process probabilistic. Because there's always less than a 100% probability that signaling from an upstream neuron will sufficiently excite a downstream neuron, there will be occasions where a downstream neuron fails to fire.

So taking these things into account, we can define the following:

  • SUCCESS = Neuron A fires -> causes Neuron B to fire
  • FAILURE = Neuron A fires -> Neuron B does not fire
An example

If a synapse has 100 channels, and each channel has a 50% chance of being open at any moment, then we expect that, on average, 50 channels will be open at any given time. If each channel can allow 1pA of current to pass through, this synapse will transmit on average 50pA of current from Neuron A to Neuron B. Let's say 25pA is needed to bring Neuron B to threshold and cause it to fire. Then this synapse will typically transmit more than enough current to evoke a response in Neuron B.

But here's the catch. Because the inactivation of each channel is spontaneous, sometimes there will be more than 50 channels open, and sometimes less than 50 - just by chance. On some rare occasions, fewer than 25 channels will be open, and if Neuron A fires at one of these crucial moments, our 100 channel synapse will transmit less than 25pA, and fail to bring Neuron B to threshold. This is how an electrical synapse "failure" occurs. You can imagine how, evolutionarily, this might be a bad thing if that failure to fire Neuron B occurs at a crucial moment when the muscle under control of Neuron B is needed to jump out of a predator's reach.

This leads us to the question: Given these definitions, how would we increase the chance of success, and correspondingly decrease the chance of failure?

Scaling up

Well, probability makes things a numbers game, and here's where size comes into play. One way we could make sure that Neuron A has a really good chance of firing Neuron B is to mitigate the risk from spontaneous channel inactivation. We could do this by simply scaling up the synapse - making it larger and adding more channels. If we had 200 synapses, rather than 100, and they all still had a 50% chance of being closed at a given moment, then the chance becomes incredibly remote that, at a given moment, fewer than 25 channels (the number needed to fire Neuron B) are open and able to transmit signal. This is what's meant by being made fail-safe.

It's easy to see that another means to the same end is to scale up the corresponding response in Neuron B - the EPSP - as observed at the neuromuscular junction. By scaling up the strength of the response, it becomes less likely that random, uncontrollable influences on the cell (or muscle) would cause a failure to fire.

Hopefully, now you can see how neurons may fail to signal, and why scaling things up can, in general, make a synapse more fail-safe.

  • $\begingroup$ It would be great if you can add some references that people can go to if they need to know the details. References are, in general, always appreciated. $\endgroup$
    Sep 21, 2016 at 8:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .