A recent study proposes that the random and spontaneous, NMDA receptor–dependent miniature postsynaptic currents at hippocampal synapses serve to regulate local postsynaptic protein synthesis, thereby stabilizing synaptic function.
Spontaneous neurotransmitter release is a common feature of synapses throughout the nervous system, but does it serve a purpose? Since Fatt and Katz discovered neurotransmission in the absence of nerve impulses1, this question has been in the minds of neurophysiologists. Spontaneous neurotransmission is widely studied, though most often as a simpler proxy for the more complicated action potential–driven synchronized release of neurotransmitters. A few studies have examined spontaneous neurotransmission for its own sake and have shown that spontaneous release events may trigger action potential firing in cells with high membrane resistance and may also be required for the maturation of synapses2. Another mysterious phenomenon that has puzzled neurophysiologists for over a century is the hyperexcitability of target membranes that follows the denervation or other disruption of their nervous input3. In the neuromuscular junction, the increase in sensitivity of muscle tissue to acetylcholine seen after denervation is due to the upregulation of acetycholine receptors4. A similar receptor upregulation occurs at central synapses, revealing a powerful mechanism for the maintenance of homeostatic stability of CNS synaptic networks5,6. Furthermore, chronic blockade of action potential firing in neuronal cultures increases trafficking of the AMPA receptor subunits GluR1 and GluR2 to postsynaptic sites, thus increasing sensitivity to released glutamate7.
A new study in Cell by Sutton and colleagues8 bridges these two persistent questions of neurophysiology through a comprehensive set of experiments and provides a causal link between the two phenomena. The authors show that spontaneous neurotransmitter release, rather than evoked neurotransmission, is a specific regulator of postsynaptic sensitivity to neurotransmitters—by suppressing the dendritic protein translation machinery locally and thereby maintaining receptor composition at synapses.
The initial set of experiments of Sutton et al. documented that, unlike the blockade of action potentials, inhibition of either NMDA receptors or AMPA receptors can increase the amplitude of miniature excitatory postsynaptic currents ('minis' or mEPSCs) within mere hours. The frequency of minis remains unchanged, and the increased mini amplitude is seen even when action potentials are allowed during receptor blockade. This finding has two surprising aspects. First, it strongly suggests that NMDA receptors are active at rest during spontaneous neurotransmission, despite their reduced ion conductance due to Mg2+ block. Second, the observation that amplitudes of unitary synaptic responses increase rapidly within an hour after NMDA receptor blockade stands in striking contrast to earlier reports that chronic blockade of neuronal firing by tetrodotoxin leads to slow rescaling of unitary synaptic efficacy6. Moreover, the authors show that this rapid effect of NMDA receptor blockade on unitary transmission is strictly dependent on protein synthesis, consistent with recent findings from the same group9.
To test whether the increase in the amplitudes of mEPSCs is due to increased cell surface expression of AMPA receptors, the authors used a biotinylation assay, and detected an increase in GluR1 but not GluR2 subunits at the cell surface. To visualize the spatial distribution of these newly recruited GluR1 subunits, they labeled surface receptors in live cultures with specific antibodies and also immunostained for synapse markers. These experiments showed that the increase in GluR1 subunits takes place at synapses. Furthermore, whereas blockade of NMDA receptors increased both the size of individual synaptic GluR1 clusters and the numbers of synapses expressing GluR1, it did not affect the expression levels of other synaptic markers. The authors also did a technically demanding set of experiments using a dual micropipette local perfusion system to inhibit NMDA receptors focally; these experiments revealed selective surface expression of GluR1 subunits near the inhibition site but not in dendritic regions away from this site. Thus, the increase of GluR1 expression after NMDA receptor block is under tight temporal and spatial regulation.
An overall increase in surface expression in GluR1 subunits but not GluR2 subunits suggests a concomitant increase of synaptic GluR2-lacking receptors, which are incidentally permeable to Ca2+ (ref. 10). To characterize the subunit composition of AMPA receptors electrophysiologically, the authors used a selective inhibitor of GluR2-lacking AMPA receptors, 1-napthylacetylspermine (Naspm). Once synapses were scaled up after the application of the NMDA receptor blocker D(-)-2-amino-5-phosphonovaleric acid (AP5), the increased amplitude of mEPSCs was decreased upon application of Naspm, suggesting a rise in the contribution of GluR2-lacking receptors after AP5 treatment. The authors also used fluorescence imaging to visualize the increase in GluR2-lacking, and thus Ca2+-permeable, AMPA receptors on dendrites. For this purpose, they took advantage of the Co2+ permeability of Ca2+-permeable AMPA receptors, a unique feature among Ca2+-permeable channels. Co2+ influx can be detected as a decrease in calcein fluorescence, as Co2+ acts as a potent quencher. These measurements confirmed the increase of GluR2-lacking AMPA receptors after prolonged focal AP5 treatment and helped the authors bridge their electrophysiology and immunocytochemistry results. Application of the protein synthesis inhibitor anisomycin together with AP5 also impaired the increase in GluR2-lacking AMPA receptors, corroborating the authors' earlier results on local protein synthesis dependence of synaptic scaling8 (Fig. 1).
Transient recruitment of GluR2-lacking AMPA receptors also occurs shortly after the induction of long-term potentiation11. GluR2-lacking AMPA receptors seem to serve as pioneers for changes in postsynaptic efficacy and may initiate the synthesis or recruitment of additional factors for maintenance of the plasticity. Thus GluR2-lacking AMPA receptors might function as a 'marker' or a 'tag' for synaptic modification. After induction of long-term potentiation as well as during homeostatic plasticity, GluR2-deficient receptors are gradually replaced with GluR2-containing ones, albeit with significantly different time courses. In a previous study, GluR2-lacking AMPA receptors were detectable only for 25 minutes11, whereas in the study of Sutton et al., the replacement of GluR2-lacking receptors by GluR2-containing ones takes place over 24 hours8. The mechanisms underlying this differential rate of receptor replacement after distinct forms of plasticity will be the interesting subject of future study.
How do AMPA and NMDA receptors cross-talk at individual synapses? The study by Sutton and colleagues, as well as earlier work12, showed that blocking AMPA receptors also induces synaptic rescaling. This finding raises the question of whether blockade of AMPA and NMDA receptors share the same downstream mechanisms. Does spontaneous release rate per synapse determine the unitary response amplitudes in a linear way? If so, does the receptor composition of a particular synapse depend on its history of spontaneous transmission? Or is spontaneous transmission merely permissive, such that above a certain threshold of spontaneous release, it can suppress translational machinery? Future identification of the signaling pathway connecting NMDA receptors to protein synthesis machinery and recruitment of GluR1 subunits may provide some answers to these questions. In addition to the cross-talk in downstream signaling, it is possible that AMPA receptor activity may augment NMDA receptor activity at rest through electrical means. Such interaction between the activation of the two types of receptors may occur if at least some dendritic spines comprise electrically isolated compartments due to high spine neck resistance13. In such circumstances, activation of AMPA receptors may result in sufficient local depolarization to facilitate the relief of adjacent NMDA receptors from Mg2+ block.
An intriguing implication of this work is the divergence of mechanisms underlying synaptic scaling after chronic inhibition of NMDA receptor–mediated miniature currents versus chronic inhibition of action potentials. The difference in the time course of action of the two maneuvers argues for separate mechanisms mediating the increase in AMPA receptor activity. One possibility is that synaptic scaling in response to chronic action potential blockade does not involve the transient expression of Ca2+-permeable AMPA receptors7. The exact mechanism mediating synaptic scaling induced by chronic tetrodotoxin treatment remains unknown. It is thought to act more globally14, which is consistent with the idea of large-scale homeostatic maintenance of synaptic circuits. If chronic blockade of action potential firing also mediates its effect synaptically but by activating a different signaling cascade, then how do neurons distinguish evoked and spontaneous neurotransmission? Are the differences in temporal characteristics of the two forms of neurotransmission sufficient? Or is there a distinct postsynaptic detection machinery for spontaneously released neurotransmitters? Spontaneous fusion events may not originate from the same pool of vesicles as evoked neurotransmission15. Better understanding of the presynaptic machinery and its postsynaptic counterparts that underlie spontaneous and evoked neurotransmission will provide us with molecular and pharmacological tools that can selectively manipulate the two forms of neurotransmission. Independent analysis of spontaneous and evoked neurotransmission may uncover more surprises in the intricacies of communication within individual synapses.
Fatt, P. & Katz, B. J. Physiol. (Lond.) 117, 109–128 (1952).
Zucker, R.S. Neuron 45, 482–484 (2005).
Cannon, W.B. & Rosenblueth, A. The Supersensitivity of Denervated Structures: A Law of Denervation (Macmillan, New York, 1949).
Axelsson, J. & Thesleff, S. J. Physiol. (Lond.) 147, 178–193 (1959).
O'Brien, R.J. et al. Neuron 21, 1067–1078 (1998).
Turrigiano, G.G., Leslie, K.R., Desai, N.S., Rutherford, L.C. & Nelson, S.B. Nature 391, 892–896 (1998).
Wierenga, C.J., Ibata, K. & Turrigiano, G.G. J. Neurosci. 25, 2895–2905 (2005).
Sutton, M.A. et al. Cell 125, 785–799 (2006).
Sutton, M.A., Wall, N.R., Aakalu, G.N. & Schuman, E.M. Science 304, 1979–1983 (2004).
Hollmann, M., Hartley, M. & Heinemann, S. Science 252, 851–853 (1991).
Plant, K. et al. Nat. Neurosci. 9, 602–604 (2006).
Thiagarajan, T.C., Lindskog, M. & Tsien, R.W. Neuron 47, 725–737 (2005).
Bloodgood, B.L. & Sabatini, B.L. Science 310, 866–869 (2005).
Turrigiano, G.G. & Nelson, S.B. Nat. Rev. Neurosci. 5, 97–107 (2004).
Sara, Y., Virmani, T., Deak, F., Liu, X. & Kavalali, E.T. Neuron 45, 563–573 (2005).
About this article
Cite this article
Chung, C., Kavalali, E. Seeking a function for spontaneous neurotransmission. Nat Neurosci 9, 989–990 (2006). https://doi.org/10.1038/nn0806-989
The Journal of Physiology (2020)
Journal of Neuroscience Research (2018)
Spontaneous neurotransmission signals through store-driven Ca2+ transients to maintain synaptic homeostasis
Differential content of vesicular glutamate transporters in subsets of vagal afferents projecting to the nucleus tractus solitarii in the rat
Journal of Comparative Neurology (2014)
Channeling satiation: A primer on the role of TRP channels in the control of glutamate release from vagal afferent neurons
Physiology & Behavior (2014)