UC San Diego

Information transfer between neurons in the central nervous system occurs at specialized contacts, called synapses. Not all excitatory (glutamatergic) synapses in the brain are identical. Rather, different inputs to the same target can have markedly different properties, as can different targets of the same presynaptic cell. Excitatory synapses exhibit long-lasting changes in efficacy after experiencing brief patterns of coordinated activity, a characteristic which is thought to underlie learning and other forms of reorganization of the nervous system. The NMDA receptor (NMDAR) serves as an exquisitely sensitive coincidence detector in the postsynaptic density, and it is crucial for spike timing dependent synaptic plasticity. However, recently, it has been proposed that in addition to the classical role, the NMDAR functions to modulate synaptic release and plasticity in novel ways. In this thesis, I first tested whether spike timing dependent plasticity in vivo follows the same rules as it does in vitro. I found evidence that the same basic rules are present (Chapter 2), but whether the phenomena studied in vitro and in vivo are the same remains to be fully confirmed. Next, I tested whether postsynaptic NMDARs exhibit calcium-dependent desensitization during coincident pre- and postsynaptic activity, as recently observed at one synapse on cortical layer 2/3 pyramidal cells. I found that this desensitization did not occur at a different input onto these neurons, suggesting that it is not a general property of NMDA receptor signaling. Finally, I investigated another novel role of NMDARs, to regulate release at presynaptic terminals. I confirmed that non-postysnaptic, likely presynaptic NMDARs regulate presynaptic release at some synapses in somatosensory cortex, but not at other synapses made by the same presynaptic cells, or made onto the same postsynaptic cells. Thus, presynaptic NMDAR function is highly target- and input-specific, indicating that presynaptic function of NMDARs underlies some of the functional heterogeneity at excitatory cortical synapses (Chapter 4). These results show that the classical view of NMDAR function is incomplete and that a full picture of synaptic release and of the diversity of excitatory synapses must include presynaptic NMDARs that govern release