IMAGE SOURCEOne of the main cellular mechanisms assumed to underlie learning is long-term potentiation (LTP), an experimental form of synaptic plasticity that results in a long-lasting increase in the strength of synaptic transmission. However, prolonged synaptic stimulation in vitro eventually stops producing further LTP (also known as 'LTP occlusion'). So, how does ongoing experience result in further learning? Clem et al. now show that the opposing actions of activated N-methyl-D-aspartate receptors (NMDARs) and metabotropic glutamate receptors (mGluRs) allow progessive synaptic strengthening during sensory-induced plasticity.
To study the link between LTP, synaptic strengthening and learning the authors used a single-whisker experience (SWE) protocol in which all but one of a mouse's whiskers were removed, resulting in the potentiation of synapses within the barrel cortical column that receives input from the remaining whisker. They demonstrated that administering an NMDAR antagonist during the initial phase of SWE prevented this synaptic strengthening. In addition, application of an NMDAR antagonist blocked the LTP that was induced by pairing (a procedure in which presynaptic stimulation is paired with postsynaptic depolarization) in barrel column synapses of control mice, indicating that NMDAR activation is required for both in vivo SWE-induced synaptic strengthening and in vitro LTP in barrel cortex synapses.
Interestingly, after 24 hours of SWE, pairing could no longer potentiate barrel column synapses — in other words, prolonged SWE caused the occlusion of further LTP at these synapses. In the past, this has been interpreted as evidence that SWE induced LTP early on and that no further LTP could be elicited as SWE continued. Crucially, however, Clem et al. showed that application of an NMDAR antagonist prevented this SWE-induced occlusion of LTP, indicating that NMDAR activation acts to suppress synaptic strengthening after prolonged sensory stimulation. Application of an mGluR antagonist plus an NMDAR antagonist to barrel column slices from SWE mice completely prevented pairing-induced LTP, suggesting that mGluR counteracts the NMDAR-mediated synaptic depression and maintains synaptic plasticity after long-term sensory experience.
The roles for NMDARs and mGluRs in synaptic plasticity in brain slices were confirmed in vivo by showing that administration of an NMDAR-antagonist to SWE mice prevented LTP occlusion, and that LTP was reduced in slices from mice that had received an mGluR antagonist. Together, these findings indicated that NMDAR activation mediates the induction of LTP at synapses of the barrel cortex, but inhibits it after prolonged stimulation, whereas activation of mGluR allows further potentiation.
Importantly, the authors also showed that NMDAR- and mGluR-mediated synaptic plasticity is linked to learning: SWE animals showed improved whisker-dependent association conditioning compared with whisker-intact mice, and treating SWE mice with an NMDAR-antagonist or an mGluR-antagonist enhanced or reduced learning, respectively.
These data provide evidence that LTP, synaptic strengthening and learning are probably underpinned by activation of NMDARs followed by activation of mGluRs. This process would explain why, despite NMDAR-dependent LTP occlusion, learning remains possible with continuing experience.