Synapse #fail, Science #win

By Fabiana Kubke 25/11/2009

ResearchBlogging.orgThe endbulb or calyx of Held is a very large synapse found in the auditory system. It consists of a very large ‘calyceal’ ending, literally wrapping around the cell body of the postsynaptic neuron. It was first described by H Held in the late 1800’s and has since been shown to characteristically be present in neuronal circuits that require very high temporal precision. (It is, by the way, my favourite synapse.)

Because the synapse is so large, there are numerous sites of contact where the neurotransmitters are released, which will happen whenever an action potential reaches the synaptic terminal. Because of this, it has always been thought that these synapses never fail to produce a response (action potential) on its target (postsynaptic) neuron, that is, that it is a fail-safe synapse: every time that there is neurotransmitter release, the postsynaptic neuron produces an action potential.

Endbulb of Held

Barn owl endbulb of Held (by Kubke)

But is this true?

Jeannette Lorteije, Silviu Rusu, Christopher Kushmerick and Gerard Borst examined precisely this, and they did so in a series of really elegant experiments in mice. They examined whether the discrepancies in the data regarding the degree of reliability at the enbulb or calyx of Held could be attributed to different methodological approaches or differences in the interpretation of the raw data. To examine this they did a series of recordings from cells in the Medial Nucleus of the Trapezoid Body (MNTB), which is part of the mammalian auditory system. The authors conclude that that there is a significant incidence of failures of transmission at this level of the system.

This is in contrast with the results reported by Bernard Englitz, Santra Tolnai, Marey Typlt, Jürgen Jost and Rüdolf Rübsamen. Here the authors recorded the failure at the endbulb of Held in the auditory cochlear nucleus AVCN and the calyx of Held in the MNTB in mongolian gerbils. They report that although failures of transmission were often found in AVCN, this was not the case in MNTB.

Synaptic structures analogous to the endbulb or calyx of Held are found in neuronal circuits that require high temporal precision. In the auditory system high temporal resolution is necessary for the measurement of interaural time differences, which in mammals are used to localize low frequency sound in the horizontal plane. Benedikt Grothe has argued that low frequency hearing appeared later in mammalian evolution, and that anatomical differences in a nucleus that receives inputs from the MNTB and is involved in the detection of interaural time differences (MSO) reflect this evolution. He argues that although MSO may have evolved to detect ITDs in low frequency hearing mammals (such as gerbils), its function may be different in higher frequency hearing mammals. On therefore wonders whether the differences in the data between the two studies may be related to adaptations associated with different temporal processing requirements in mammals with different frequency hearing ranges.

What did Lorteije and collaborators do?

In order to decide whether there are times in which synaptic release fails to elicit an action potential on the target cell, one needs to simultaneously monitor the activity happening at the synapse as well as at the postsynaptic neuron. There are traditionally two ways of doing this: One is to record the currents near the synapse that are produced by the electrical activity of the synapse and the cell, and the endbulbs of Held are large enough to produce sufficient current that can be detected. The other is to actually record the activity simultaneously from the cell and the synaptic terminal, which is usually done in an ‘in vitro’ preparation.

Lorteije and colleagues produced a set of data that is simply amazing, and their findings explain many of the discrepancies that can be found in the literature. They answered some very straightforward questions:

  1. Are the extracellular recordings done in vivo representative of what is actually going at a single endbulb-neuron contact? (the answer is yes)
  2. Is there synaptic  release that fails to produce an action potential in the postsynaptic neuron? (the answer is also yes)
  3. Is the short term synaptic depression seen in vitro also seen in the whole animal (in vivo)? (Short term depression is a reduction in the effect of synaptic release on the postsynaptic cell.). (The answer is basically no)

The authors recorded from cells in the Medial Nucleus of the Trapezoid Body (MNTB), which receives inputs in the form of the large calyces of Held and is involved in auditory processing. They did this by recording the spontaneous and auditory-evoked activity extracellularly (as most people do) as well as directly from the cells with a patch pipette in anaesthetized mice. They then repeated these experiments in vitro, this time simultaneously recording extracellularly and in whole cell patch, which allowed them to confirm that the extracellular recordings in vivo did indeed represent the activities of the terminal and the cell and that it could also provide information as to the size of the synaptic potential. Their results have two important findings:

  1. in vivo there is no observable short term synaptic depression. The synaptic depression observed in vitro may be partly due to the concentration of Calcium in the bathing solution, but other factors may be involved.
  2. They also found that the release of neurotransmitter at the synapse often failed to produce an action potential in the postsynaptic cell. A similar rate of failure to that observed in vivo can be obtained in vitro by lowering the calcium concentration of the bathing solution.

The authors summarize their findings by saying:

’Due to its low release probability and large number of release sites, its average output can be kept constant, regardless of firing frequency. Its low quantal output thus allows it to be a tonic synapse, but the price it pays is an increase in jitter and synaptic latency and occasional postsynaptic failures.’

This is a carefully designed study, and despite my concerns as to whether their results are generalizable to other mammals, they do provide data that will be welcome by many auditory neurophysiologists. Their ability to record from a patch in vivo is no small feat, and the correlation between intracellular and extracellular data is extremely useful. Further, there is a cautionary tale around the way that data obtained from in vitro data can be interpreted.

And if you think this post is long, try reading the paper! (There are heaps more gems in there.)


Lorteije, J., Rusu, S., Kushmerick, C., & Borst, J. (2009). Reliability and Precision of the Mouse Calyx of Held Synapse Journal of Neuroscience, 29 (44), 13770-13784 DOI: 10.1523/JNEUROSCI.3285-09.2009
Englitz, B., Tolnai, S., Typlt, M., Jost, J., & Rübsamen, R. (2009). Reliability of Synaptic Transmission at the Synapses of Held In Vivo under Acoustic Stimulation PLoS ONE, 4 (10) DOI: 10.1371/journal.pone.0007014
Grothe, B. (2000). The evolution of temporal processing in the medial superior olive, an auditory brainstem structure Progress in Neurobiology, 61 (6), 581-610 DOI: 10.1016/S0301-0082(99)00068-4