Chronic blockade of excitatory glutamatergic synaptic receptors in co-cultured organotypic rodent neocortex explants leads to a compensatory up-regulation of otherwise inactive input channels so as to maintain almost normal levels of ongoing bursts of action potentials. We report here that this homeostatic return of spontaneous, now kainate receptor driven, firing is accompanied by a reciprocal down-regulation of the blocked AMPA and NMDA receptors, such that the developing cortical network is protected from becoming hyperactive when these synaptic inputs again become able to transmit normally.Intrinsically generated bioelectric currents, periodically triggering polyneuronal bursts and isolated spikes, is a well-nigh universal feature of neural networks, especially during early ontogeny [1]. Spontaneous burst activity (SBA) is characteristically organized as nested clusters of action potentials on increasingly higher order time-scales, which show a strong capacity for reacting homeostatically to experimental interference [2]. Thus, NMDA receptor blockade in developing organotypic rodent visual cortex cultures begins within hours to be compensated for by steadily rising firing rates, which regain stable control levels within 24h [1,2]. Although action potential discharges are then purely AMPA receptor driven, 2-weekold chronically treated cultures show no augmentation of spontaneous activity, and continue to burst normally, upon release of their NMDA receptors from the pharmacological blockade.Combined blockade of NMDA and AMPA receptors eliminates SBA throughout the treatment period in isolated cortex cultures [1] but fails to do so in co-cultured explants which have been enabled to cross-innervate one another [3]. In such preparations, kainate receptormediated SBA takes over for the blocked excitatory receptors [2], as could be shown by acute application of the highly selective kainate blocker LY382284 (courtesy of Eli Lilly & Co). Spontaneous firing levels and burst intensities failed to be fully restored, however, primarily owing to a dearth of exceptionally active explants in the experimental group (see table 1: 'growth medium'). When such cultures were subsequently assayed in control medium the treated explants became more active (table 1) but, here too, did not exceed the values recorded in the control group. Although spontaneous bursts tended to be somewhat longer than normal, they were correspondingly less frequent and intense [4]. Three-week-old cultures gave comparable results, on the whole, but were much more individually variable.One might have expected that even if, by virtue of the homeostatic restoration of SBA, blocked glutamate receptors had been prevented from becoming sensitized (and inhibitory interneurons from declining in number and/or efficacy) during the treatment period, they still ought to have contributed to intensified spontaneous firing when acting in synergy with the up-regulated kainate receptors. Their failure to do so suggests that, in accordance with a 'Hebbian' weakening of ina...