Neocortical Layer-5 tLTD Relies on Non-Ionotropic Presynaptic NMDA Receptor Signaling

In the textbook view, NMDA receptors (NMDARs) act as coincidence detectors
in Hebbian plasticity by fluxing Ca2+ when simultaneously depolarized and
glutamate bound. Hebbian coincidence detection requires that NMDARs be
located postsynaptically, but enigmatic presynaptic NMDARs (preNMDARs)
also exist. It is known that preNMDARs regulate neurotransmitter release,
but precisely how remains poorly understood. Emerging evidence suggest
that NMDARs can also signal non-ionotropically, without the need for Ca2+
flux. At synapses between developing visual cortex layer-5 (L5) pyramidal
cells (PCs), preNMDARs rely on Mg2+ and Rab3-interacting molecule 1αβ
(RIM1αβ) to regulate evoked release during periods of high-frequency
firing, but they signal non-ionotropically via c-Jun N-terminal kinase 2
(JNK2) to regulate spontaneous release regardless of frequency. At the
same synapses, timing-dependent long-term depression (tLTD) depends on
preNMDARs but not on frequency. We therefore tested if tLTD relies on
non-ionotropic preNMDAR signaling. We found that tLTD at L5 PC→PC synapses
was abolished by pre- but not postsynaptic NMDAR deletion, cementing the
view that tLTD requires preNMDARs. In agreement with non-ionotropic NMDAR
signaling, tLTD prevailed after channel pore blockade with MK-801, unlike
tLTP. Homozygous RIM1αβ deletion did not affect tLTD, but wash-in of the
JNK2 blocker SP600125 abolished tLTD. Consistent with a presynaptic need
for JNK2, a peptide blocking the interaction between JNK2 and Syntaxin-1a
(STX1a) abolished tLTD if loaded pre- but not postsynaptically, regardless
of frequency. Finally, low-frequency tLTD was not blocked by the channel
pore blocker MK-801, nor by 7-CK, a non-competitive NMDAR antagonist at
the co-agonist site. We conclude that neocortical L5 PC→PC tLTD relies on
non-ionotropic preNMDAR signaling via JNK2/STX1a. Our study brings closure
to long-standing controversy surrounding preNMDARs and highlights how the
textbook view of NMDARs as ionotropic coincidence detectors in plasticity
needs to be reassessed.