RHYTHMICAL BURSTS INDUCED BY NMDA IN GUINEA-PIG CHOLINERGIC NUCLEUS BASALIS NEURONS IN-VITRO

1. Intracellular recordings were performed in neurones within the basa l forebrain of guinea-pig brain slices. Following injection of biocyti n (or biotinamide), a subset of recorded neurones which displayed dist inct intrinsic membrane properties were confirmed as being cholinergic by immuno-histochemical staining for choline acetyltransferase (ChAT) . They were all located within the nucleus basalis magnocellularis. Th e response of the cholinergic cells to NMDA and to the agonists of the other glutamate receptors was tested by bath application of NMDA, t-A CPD, AMPA and kainate. 2. When depolarized from a hyperpolarized level , cholinergic basalis neurones display the intrinsic ability to discha rge in rhythmic bursts that are generated by low-threshold Ca2+ spikes . In control solution, these rhythmic bursts were not sustained for mo re than 5-6 cycles. However, in the presence of NMDA when the membrane was held at it hyperpolarized level, low-threshold bursting activity was sustained for prolonged periods of time. This activity could be re versibly eliminated by D(-)-2-amino-5-phosphonopentanoic acid (D-AP5), showing that it depended upon specific activation of NMDA receptors. 3. NMDA-induced, voltage-dependent, rhythmic depolarizations persisted in the presence of tetrodo-toxin (TTX), indicating that they did not depend upon a TTX-sensitive Na+ current and were generated postsynapti cally. The rhythmic depolarizations mere, however, eliminated by tile par tial replacement of of Na+ with choline, demonstrating that they d id depend upon Na+, the major carrier of the NMDA current. 4. In the p resence of TTX, the NMDA-induced rhythmic depolarizations were also el iminated by removal of Ca2+ from or addition of Ni2+ to the bath, indi cating that they also depended upon Ca2+, which is carried by both the NMDA current and the low-threshold Ca2+ current. The duration of the rhythmic depolarizations was increased in the presence of apamin, sugg esting that the repolarization of the cells depended in part upon a Ca 2+ activated K+ (SK) conductance, but that other mechanisms were addit ionally involved in the repolarization phase of the bursting. 5. In bo th the absence and presence of TTX, the NMDA-induced rhythmic activity persisted when Mg2+ was removed from the medium, indicating that the sustained rhythmic depolarizations did not hinge upon the Mg2+ block o f the NMDA channels during hyperpolarization. The voltage dependence o f the NMDA-induced rhythmic depolarizations in the absence of Mg2+ app eared to be determined by the properties of the low-threshold Ca2+ spi ke in the cholinergic basalis neurones. 6. These in vitro results show that activation of NMDA channels excites cholinergic basalis neurones and mag drive them into tonic firing if allowed to depolarize fully o r maintain them in a rhythmic bursting mode if they are simultaneously held at a hyperpolarized level from which intrinsic lon thr threshold Ca2+ spikes are triggered. Assuming the presence of contingent hyperp olarizing afferent input, these data suggest that brainstem and cortic al afferents that release glutamate could stimulate rhythmic bursting sia NMDA receptors in the cholinergic cells ill vivo. Such rhythmic os cillations in the basalis neurones would provide a rhythmic modulation to target neurones within the cerebral cortex and thereby potentially promote slow oscillations within a delta or theta frequency range in cortical activity across the sleep-waking cycle.