Distinct modifications of convergent excitatory and inhibitory inputs in developing olfactory circuits

The interaction between excitatory and inhibitory inputs is critical to neuronal signal processing. However, little is known about this fundamental property, largely due to the inability to clearly isolate the respective inputs. Here we took advantage of the characteristic stereotypical architecture of synaptic connections in the main olfactory bulb, which enabled us to entirely separate excitatory and inhibitory inputs. Using paired stimulation of two glomeruli located apart at different intensities, we separately elicited excitatory and inhibitory inputs and mimicked stimulation of competing mitral cells (MCs) with different odorants. We performed dual whole-cell patch recording of evoked excitatory postsynaptic responses (EPSPs) and inhibitory postsynaptic responses (IPSPs) in current-clamp mode from two competitive MCs that are connected to the two stimulated glomeruli in slices of the main olfactory bulb in 2-3-week-old rats. We deliberately held the recorded cells at a relative hyperpolarized potential. This manipulation not only suppressed action potential generation but also excluded the possible contamination of inhibitory components in excitatory inputs. We found that in weakly activated MCs repetitive EPSP-IPSP interactions (5 Hz, 180 times) induced long-term potentiation (LTP) and long-term depression (LTD) in convergent excitatory and inhibitory inputs, respectively. Unexpectedly, these forms of plasticity depend on activity of somatic (mainly non-synaptic) NMDA receptors (NMDARs). In contrast, the same repetitive stimulation induced the LTP of excitatory inputs in strongly activated MCs (MC2) that require activity of synaptic NMDARs. These distinct forms of plasticity in the developing olfactory circuit may represent a novel rule of modification in convergent inputs that leads to decorrelation of inputs and facilitates odor discrimination.