From neural plate to cortical arousal-a neuronal network theory of sleep derived from in vitro "model" systems for primordial patterns of spontaneous bioelectric activity in the vertebrate central nervous system

Abstract

In the early 1960s intrinsically generated widespread neuronal discharges were discovered to be the basis for the earliest motor behavior throughout the animal kingdom. The pattern generating system is in fact programmed into the developing nervous system, in a regionally specific manner, already at the early neural plate stage. Such rhythmically modulated phasic bursts were next discovered to be a general feature of developing neural networks and, largely on the basis of experimental interventions in cultured neural tissues, to contribute significantly to their morpho-physiological maturation. In particular, the level of spontaneous synchronized bursting is homeostatically regulated, and has the effect of constraining the development of excessive network excitability. After birth or hatching, this "slow-wave" activity pattern becomes sporadically suppressed in favor of sensory oriented "waking" behaviors better adapted to dealing with environmental contingencies. It nevertheless reappears periodically as "sleep" at several species-specific points in the diurnal/nocturnal cycle. Although this "default" behavior pattern evolves with development, its essential features are preserved throughout the life cycle, and are based upon a few simple mechanisms which can be both experimentally demonstrated and simulated by computer modeling. In contrast, a late onto- and phylogenetic aspect of sleep, viz., the intermittent "paradoxical" activation of the forebrain so as to mimic waking activity, is much less well understood as regards its contribution to brain development. Some recent findings dealing with this question by means of cholinergically induced "aroused" firing patterns in developing neocortical cell cultures, followed by quantitative electrophysiological assays of immediate and longterm sequelae, will be discussed in connection with their putative implications for sleep ontogeny.

Figure 1

Sketch made for a lecture in 1961 at the Zoology department, Columbia University (New York City) showing: left—the limb/nerve deplantation experiment of Paul Weiss in the axolotl dorsal fin; middle—a frog tadpole showing the transections made at different CNS levels; right—outline of the early anuran neural plate illustrating the spontaneous motility patterns generated by isolated presumptive CNS fragments combined with mesoderm (see text and [6] for details).

Figure 2

Two examples of “Frankensteinian” preparations prepared and filmed in 1961 at the Hubrecht International Embryology Laboratory in Utrecht, The Netherlands, consisting of anuran neural plate tissue, plus mesoderm, encased in a transparent ball of ectoderm. Differentiated CNS fragments (N) are covered by a pigment layer, and twitches of the muscle fibers (M) could be easily followed through a stereo microscope.

Figure 3

Oscillogram photographed in 1963 at Columbia University, Department of Neurology, showing the first ever example of a spontaneous polyneuronal barrage of field and action potentials (top trace) driving a stereotyped burst of muscle twitches. Recorded in an organotypic culture derived from a presumptive brainstem motor region of the frog neural plate. Time bar = 1 s.

Figure 4

Oscillogram (slightly retouched) of stereotyped polyneuronal burst firing (upper trace) as well as continuous field potentials associated with tonic firing during an exceptionally active period in an organotypic mouse medulla culture. Recorded in 1968 at the Rose Kennedy Center, Albert Einstein College of Medicine, New York, NY, USA.

Figure 5

Examples of all-or-none responses triggered in different preparations or at different recording sites by electrical stimulation in organotypic mouse medulla cultures. Recorded in 1968 at the Rose Kennedy Center, Albert Einstein College of Medicine, New York, NY, USA.

Figure 6

Different patterns of spontaneous field potentials or isolated spiking (A–C), and all-or-none evoked responses (D–F) in organotypic chick embryo cerebral hemisphere explants: oscillographically recorded in 1968 at the Albert Einstein College of Medicine, New York, NY, USA.

Figure 7

Examples of (four) different classes of more or less stereotyped neuronal firing in a dissociated rat neocortex culture recorded on a multi-electrode plate. The connecting lines indicate the sequential firing of individual spikes at the participating recording sites: experiment performed together with Dr. van Pelt, J. in 2005 at The Netherlands Institute for Brain Research, Amsterdam, The Netherlands.

Figure 8

Comparison between control cultures (blue: n = 6) and overnight TTX (20 μM) pre-treated cultures (red: n = 7) for selected parameters of spontaneous activity. MFR = mean firing rate; burst intensity = intraburst firing rate; inverse ratio = the proportion of spikes falling outside bursts (interspike interval detection criterion for bursts ≤100 ms). All values were normalized with respect to the mean level in the 2-h period prior to administering 20 μM carbachol (for ~20 h), followed by ~20 h of washout. The individual means at each time point were then used to calculate a grand mean for each group, with error bars indicating the SEM. (For examples of actual polyneuronal firing patterns throughout a representative experiment, see [13]).

Figure 9

The same as in Figure 8 for two additional parameters descriing spontaneous neuronal firing in dissociated neocortical cell cultures: left—the estimated overall excitability relative to the pre-carbachol baseline is given by the “Single Pulse Response” parameter (see [90] for details); right—the Burst Index (BI: see [63] shows a strong and prolonged post-carbachol (20–40 h) augmentation in both groups, along with a very large inter-individual variance in the TTX pre-treated group.