Disinhibition of MCs by inhibiting periglomerular neurons could enhance MC excitation and permit the level of calcium entry required for synaptic plasticity. Our data show that enhancing glomerular inhibition by muscimol puffs to the glomerular layer prevents the potentiation of MC spikes induced by the pairing of TBS and isoproterenol. Furthermore, local, as indicated by phenol red, glomerular disinhibition by a low dose of gabazine, when paired with TBS, results in MC spike potentiation in the absence of isoproterenol. Further tests of the effect of isoproterenol on directly connected PG and MCs using paired electrical recording or optical imaging would provide more direct evidence regarding the role of isoproterenol and periglomerular disinhibtion. Given these effects on ON-MC plasticity, we employed an infusion method to study the effect of glomerular disinhibition on learning. We reasoned that if isoproterenol causes depolarization of MCs through glomerular disinhibition, then enhancing glomerular inhibition by muscimol infusion would prevent isoproterenol-mediated early odor preference learning. Our results show that co-infusion of muscimol completely blocked isoproterenol- induced learning; however this could have occurred due to loss of odor signaling at encoding. We tested this possibility by examining the effect of muscimol on normal peppermint aversions. The same infusion of muscimol prevented the normal peppermint aversion seen without training and suggests that odor signaling is altered by muscimol. More convincingly in support of an important role for disinhibition, our data showed that a glomerular infusion of the GABAA antagonist, gabazine, paired with odor, produced an odor preference. We infer that local disinhibition in the glomeruli responding to peppermint is sufficient for odor preference learning. We cannot, of course, rule out in the in vivo infusions that there is some contribution of granule cell disinhibition. The work of Kaba demonstrated that manipulation of inhibition in the olfactory bulb by whole bulbar infusion of a GABAA receptor agonist, or antagonist, blocked or induced odor learning in PND 12 rats. In our preparation, the use of a lateral infusion method, which did not increase MC excitation as indicated in our pCREB test in the D-APV infusion experiment, suggests glomerular disinhibition alone may be effective in inducing learning. This is supported by the in vitro plasticity data using PI-103 distributor gabazine puff application to the glomeruli. We propose that both beta-adrenergic-mediated disinhibition and phosphorylation of mitral cell GluN1 subunits act in concert to enhance NMDA calcium currents and promote local olfactory OTX015 nerve-mitral cell potentiation. While the NMDAR is activated during learning acquisition, we found a down-regulation of NMDAR responses during the memory phase.