Genetic elimination of GABAergic neurotransmission reveals two distinct pacemakers for spontaneous waves of activity in the developing mouse cortex.

Many structures of the mammalian CNS generate propagating waves of electrical activity early in development. These waves are essential to CNS development, mediating a variety of developmental processes, such as axonal outgrowth and pathfinding, synaptogenesis, and the maturation of ion channel and receptor properties. In the mouse cerebral cortex, waves of activity occur between embryonic day 18 and postnatal day 8 and originate in pacemaker circuits in the septal nucleus and the piriform cortex. Here we show that genetic knock-out of the major synthetic enzyme for GABA, GAD67, selectively eliminates the picrotoxin-sensitive fraction of these waves. The waves that remain in the GAD67 knock-out have a much higher probability of propagating into the dorsal neocortex, as do the picrotoxin-resistant fraction of waves in controls. Field potential recordings at the point of wave initiation reveal different electrical signatures for GABAergic and glutamatergic waves. These data indicate that: (1) there are separate GABAergic and glutamatergic pacemaker circuits within the piriform cortex, each of which can initiate waves of activity; (2) the glutamatergic pacemaker initiates waves that preferentially propagate into the neocortex; and (3) the initial appearance of the glutamatergic pacemaker does not require preceding GABAergic waves. In the absence of GAD67, the electrical activity underlying glutamatergic waves shows greatly increased tendency to burst, indicating that GABAergic inputs inhibit the glutamatergic pacemaker, even at stages when GABAergic pacemaker circuitry can itself initiate waves.

Large-scale waves of activity in the neonatal mouse brain in vivo occur almost exclusively during sleep cycles.

Spontaneous electrical activity plays major roles in the development of cortical circuitry. This activity can occur highly localized regions or can propagate over the entire cortex. Both types of activity coexist during early development. To investigate how different forms of spontaneous activity might be temporally segregated, we used wide-field trans-cranial calcium imaging over an entire hemisphere in P1-P8 mouse pups. We found that spontaneous waves of activity that propagate to cover the majority of the cortex (large-scale waves; LSWs) are generated at the end of the first postnatal week, along with several other forms of more localized activity. We further found that LSWs are segregated into sleep cycles. In contrast, cortical activity during wake states is more spatially restricted and the few large-scale forms of activity that occur during wake can be distinguished from LSWs in sleep based on their initiation in the motor cortex and their correlation with body movements. This change in functional cortical circuitry to a state that is permissive for large-scale activity may temporally segregate different forms of activity during critical stages when activity-dependent circuit development occurs over many spatial scales. Our data also suggest that LSWs in early development may be a functional precursor to slow sleep waves in the adult, which play critical roles in memory consolidation and synaptic rescaling.

Developmental changes in propagation patterns and transmitter dependence of waves of spontaneous activity in the mouse cerebral cortex.

Waves of spontaneous electrical activity propagate across many regions of the central nervous system during specific stages of early development. The patterns of wave propagation are critical in the activation of many activity-dependent developmental programs. It is not known how the mechanisms that initiate and propagate spontaneous waves operate during periods in which major changes in neuronal structure and function are taking place. We have recently reported that spontaneous waves of activity propagate across the neonatal mouse cerebral cortex and that these waves are initiated at pacemaker sites in the septal nucleus and ventral cortex. Here we show that spontaneous waves occur between embryonic day 18 (E18) and postnatal day 12 (P12), and that during that period they undergo major changes in transmitter dependence and propagation patterns. At early stages, spontaneous waves are largely GABA dependent and are mostly confined to the septum and ventral cortex. As development proceeds, wave initiation depends increasingly on AMPA-type glutamate receptors, and an ever increasing fraction of waves propagate into the dorsal cortex. The initiation sites and restricted propagation of waves at early stages are highly correlated with the position of GABAergic neurons in the cortex. The later switch to a glutamate-based mechanism allows propagation of waves into the dorsal cortex, and appears to be a compensatory mechanism that ensures continued wave generation even as GABA transmission becomes inhibitory.

Early network activity propagates bidirectionally between hippocampus and cortex.

Spontaneous activity in the developing brain helps refine neuronal connections before the arrival of sensory-driven neuronal activity. In mouse neocortex during the first postnatal week, waves of spontaneous activity originating from pacemaker regions in the septal nucleus and piriform cortex propagate through the neocortex. Using high-speed Ca(2+) imaging to resolve the spatiotemporal dynamics of wave propagation in parasagittal mouse brain slices, we show that the hippocampus can act as an additional source of neocortical waves. Some waves that originate in the hippocampus remain restricted to that structure, while others pause at the hippocampus-neocortex boundary and then propagate into the neocortex. Blocking GABAergic neurotransmission decreases the likelihood of wave propagation into neocortex, whereas blocking glutamatergic neurotransmission eliminates spontaneous and evoked hippocampal waves. A subset of hippocampal and cortical waves trigger Ca(2+) waves in astrocytic networks after a brief delay. Hippocampal waves accompanied by Ca(2+) elevation in astrocytes are more likely to propagate into the neocortex. Finally, we show that two structures in our preparation that initiate waves-the hippocampus and the piriform cortex-can be electrically stimulated to initiate propagating waves at lower thresholds than the neocortex, indicating that the intrinsic circuit properties of those regions are responsible for their pacemaker function.

Bilaterally propagating waves of spontaneous activity arising from discrete pacemakers in the neonatal mouse cerebral cortex.

Spontaneous electrical activity that moves in synchronized waves across large populations of neurons plays widespread and important roles in nervous system development. The propagation patterns of such waves can encode the spatial location of neurons to their downstream targets and strengthen synaptic connections in coherent spatial patterns. Such waves can arise as an emergent property of mutually excitatory neural networks, or can be driven by a discrete pacemaker. In the mouse cerebral cortex, spontaneous synchronized activity occurs for approximately 72 h of development centered on the day of birth. It is not known whether this activity is driven by a discrete pacemaker or occurs as an emergent network property. Here we show that this activity propagates as a wave that is initiated at either of two homologous pacemakers in the temporal region, and then propagates rapidly across both sides of the brain. When these regions of origin are surgically isolated, waves do not occur. Therefore, this cortical spontaneous activity is a bilateral wave that originates from a discrete subset of pacemaker neurons.

Properties and mechanisms of spontaneous activity in the embryonic chick hindbrain.

Spontaneous activity regulates many aspects of central nervous system development. We demonstrate that in the embryonic chick hindbrain, spontaneous activity is expressed between embryonic days (E) 6-9. Over this period the frequency of activity decreases significantly, although the events maintain a consistent rhythm on the timescale of minutes. At E6, the activity is pharmacologically dependent on serotonin, nACh, GABA(A), and glycine input, but not on muscarinic, glutamatergic, or GABA(B) receptor activation. It also depends on gap junctions, t-type calcium channels and TTX-sensitive ion channels. In intact spinal cord-hindbrain preparations, E6 spontaneous events originate in the spinal cord and propagate into lateral hindbrain tissue; midline activity follows the appearance of lateral activity. However, the spinal cord is not required for hindbrain activity. There are two invariant points of origin of activity along the midline, both within the caudal group of serotonin-expressing cell bodies; one point is caudal to the nV exit point while the other is caudal to the nVII exit point. Additional caudal midline points of origin are seen in a minority of cases. Using immunohistochemistry, we show robust differentiation of the serotonergic raphe near the midline at E6, and extensive fiber tracts expressing GAD65/67 and the nAChR in lateral areas; this suggests that the medial activity is dependent on serotonergic neuron activation, while lateral activity requires other transmitters. Although there are differences between species, this activity is highly conserved between mouse and chick, suggesting that developmental event(s) within the hindbrain are dependent on expression of this spontaneous activity.

Roles of glutamate and GABA receptors in setting the developmental timing of spontaneous synchronized activity in the developing mouse cortex.

Spontaneous, synchronized electrical activity (SSA) plays important roles in nervous system development, but it is not clear what causes it to start and stop at the appropriate times. In previous work, we showed that when SSA in neonatal mouse cortex is blocked by TTX in cultured slices during its normal time of occurrence (E17-P3), it fails to stop at P3 as it does in control cultured slices, but instead persists through at least P10. This indicates that SSA is self-extinguishing. Here we use whole-cell recordings and [Ca2+]i imaging to compare control and TTX-treated slices to isolate the factors that normally extinguish SSA on schedule. In TTX-treated slices, SSA bursts average 4 s in duration, and have two components. The first, lasting about 1 s, is mediated by AMPA receptors; the second, which extends the burst to 4 s and is responsible for most of the action potential generation during the burst, is mediated by NMDA receptors. In later stage (P5-P9) control slices, after SSA has declined to about 4% of its peak frequency, bursts lack this long NMDA component. Blocking this NMDA component in P5-P9 TTX-treated slices reduces SSA frequency, but not to the low values found in control slices, implying that additional factors help extinguish SSA. GABA(A) inhibitors restore SSA in control slices, indicating that the emergence of GABA(A)-mediated inhibition is another major factor that helps terminate SSA.