NMDA receptor‐independent long‐term potentiation in hippocampal interneurons

Long‐term potentiation (LTP) of excitatory synaptic transmission mediated by α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazole‐propionic acid (AMPA) receptors in the hippocampus has been regarded as a neural basis of memory formation. In the CA1 region of the hippocampus, LTP is induced by activation of postsynaptic N‐methyl‐d‐aspartate (NMDA) receptors usually caused by high‐frequency firings of presynaptic afferent fibres (Schaffer collaterals) originating from CA3 pyramidal cells. Therefore, the induction of LTP in this region requires simultaneous presynaptic and postsynaptic activities, and LTP is typically expressed in the specific synapses activated by stimulated afferent fibres, which is called ‘synaptic specificity’.

One of the characteristic properties of NMDA receptors, which is not usually observed in AMPA receptors, is their permeability to Ca²⁺ and the increase in Ca²⁺ concentrations mediated by NMDA receptors activates postsynaptic Ca²⁺‐dependent kinases including Ca²⁺/calmodulin‐dependent protein kinase II (CaMKII) especially localized in spines on the dendrite. Activation of the kinases results in the insertion of AMPA receptors to the membrane of spines, which causes long‐lasting enhancement of excitatory synaptic transmission. This is the most generally accepted mechanism of LTP in the pyramidal cell of the CA1 region of the hippocampus; however, the increase in Ca²⁺ in postsynaptic cells is also mediated by other sources, including Ca²⁺ channels on the postsynaptic membrane and intracellular Ca²⁺ stores such as the endoplasmic reticulum and mitochondria.

It has been shown that repetitive depolarizing pulses applied to the postsynaptic cell induce NMDA receptor‐independent short‐term potentiation of excitatory synaptic transmission in the pyramidal cell of the hippocampal CA1 region (Kullmann et al. 1992). This transient potentiation requires the rise of postsynaptic Ca²⁺ concentrations, which is mediated by the activation of postsynaptic L‐type Ca²⁺ channels. Because the size of miniature excitatory postsynaptic currents (mEPSCs), which are the smallest synaptic response at a synapse and a measure of postsynaptic sensitivity to glutamate, is increased during the potentiation to the same level of the potentiation observed in evoked synaptic responses, depolarizing pulses increase the postsynaptic sensitivity not in a synapse‐specific but in a neuron‐wide manner. In some condition, depolarizing pulses induce NMDA receptor‐independent LTP instead of short‐term potentiation, which shares the expression mechanism with NMDA receptor‐dependent LTP induced by high‐frequency tetanic stimulation of afferent fibres (Kato et al. 2009). Further analyses have revealed that the metabotropic glutamate receptor 5 (mGluR5) and the InsP₃ receptor are involved in this type of LTP (Kato et al. 2012).

Although the properties of depolarizing pulse‐induced potentiation have been extensively examined and key molecules associated with the potentiation have been identified in CA1 pyramidal cells, it has not been well known whether similar phenomena are observed in interneurons. In this issue of The Journal of Physiology, Nicholson and Kullmann (2017) demonstrate that the T‐type Ca²⁺ channel plays a pivotal role in the induction of a similar type of LTP in hippocampal stratum oriens‐alveus (O/A) interneurons. In a previous paper by the same group (Le Duigou & Kullmann, 2011), exogenous activation of group I mGluRs with postsynaptic hyperpolarization was shown to be sufficient to induce LTP in the absence of presynaptic activities in hippocampal interneurons. This finding is rather unexpected because it is unclear why hyperpolarization in addition to mGluR activation is required for this type of LTP.

In the present study, Nicholson and Kullmann (2017) clearly showed that NMDA receptor‐independent LTP induced by high‐frequency stimulation of afferent fibres in the presence of the NMDA receptor antagonist d‐2‐amino‐5‐phosphonovaleric acid (d‐APV) in O/A interneurons paired with postsynaptic hyperpolarization was almost completely inhibited by T‐type Ca²⁺ channel blockers, indicating that the Ca²⁺ influx through T‐type Ca²⁺ channels is another source of Ca²⁺ required for LTP induction. It is known whether this type of Ca²⁺ channels is abundantly expressed in hippocampal interneurons, and in fact, the authors explicitly demonstrated electrophysiologically that T‐type Ca²⁺ channels were functional in O/A interneurons. They also found that non‐associative LTP caused by repetitive firings in the hyperpolarized postsynaptic cell was reduced by T‐type Ca²⁺ channel blockers in O/A interneurons. Interestingly, LTP induced by the application of the group I mGluR agonist (S)‐3,5‐dihydroxyphenylglycine (DHPG) in combination with postsynaptic hyperpolarizing currents was also completely blocked by T‐type Ca²⁺ channel blockers.

These lines of evidence strongly suggest that T‐type Ca²⁺ channels are involved in the induction of this type of LTP in O/A interneurons of the hippocampus. The T‐type Ca²⁺ channel is known to be inactivated even by a small depolarization and a substantial population of the T‐type Ca²⁺ channels is inactivated at the resting membrane potential. Thus, the authors concluded that the hyperpolarization required for this type of synaptic potentiation reduced the inactivation of T‐type Ca²⁺ channels, which enabled full activation of the channels and resulted in a sufficient increase of postsynaptic Ca²⁺ concentrations necessary for LTP induction.

Although the present study provides novel and interesting findings, it also raises some important questions. Firstly, it seems unclear how T‐type Ca²⁺ channels are activated when the postsynaptic cell is forced to be hyperpolarized by current injection.

Tetanic stimulation of afferent fibres or action potentials in the postsynaptic cell may cause some level of depolarization that is sufficiently fast to activate T‐type Ca²⁺ channels, but mGluRs may not be able to depolarize the cell sufficiently and even if they depolarize the cell, the depolarization should occur very slowly and T‐type Ca²⁺ channels must be inactivated. It should be noted, however, that spontaneous excitatory synaptic responses may cause sufficient depolarization in the spine to activate T‐type Ca²⁺ channels. It is also plausible that the membrane potential in the dendrite is not the same as that in the soma, which enables sufficient activation of T‐type Ca²⁺ channels on the spine and dendrite. Secondly, it would be interesting to know how O/A interneurons are hyperpolarized in more physiological conditions. In the present study, hyperpolarization was artificially caused by current injection, but such large hyperpolarization occurs in quite particular conditions in the brain. Strong inhibitory synaptic inputs from many other interneurons that simultaneously fire or a large afterhyperpolarization that appears after repetitive firings may be a candidate. In any case, it is intriguing to know the exact mechanism of the induction of this type of LTP and whether this phenomenon is observed in more physiological conditions and in the brain. Finally, it is important to explore the physiological significance of this type of LTP. For instance, to examine the effect of this synaptic potentiation on LTP at excitatory synapses would be an interesting issue in future studies.

Additional information

Competing interests

None declared.