Dynamic Hebbian Cross-Correlation Learning Resolves the Spike Timing Dependent Plasticity Conundrum.

Spike Timing-Dependent Plasticity has been found to assume many different forms. The classic STDP curve, with one potentiating and one depressing window, is only one of many possible curves that describe synaptic learning using the STDP mechanism. It has been shown experimentally that STDP curves may contain multiple LTP and LTD windows of variable width, and even inverted windows. The underlying STDP mechanism that is capable of producing such an extensive, and apparently incompatible, range of learning curves is still under investigation. In this paper, it is shown that STDP originates from a combination of two dynamic Hebbian cross-correlations of local activity at the synapse. The correlation of the presynaptic activity with the local postsynaptic activity is a robust and reliable indicator of the discrepancy between the presynaptic neuron and the postsynaptic neuron's activity. The second correlation is between the local postsynaptic activity with dendritic activity which is a good indicator of matching local synaptic and dendritic activity. We show that this simple time-independent learning rule can give rise to many forms of the STDP learning curve. The rule regulates synaptic strength without the need for spike matching or other supervisory learning mechanisms. Local differences in dendritic activity at the synapse greatly affect the cross-correlation difference which determines the relative contributions of different neural activity sources. Dendritic activity due to nearby synapses, action potentials, both forward and back-propagating, as well as inhibitory synapses will dynamically modify the local activity at the synapse, and the resulting STDP learning rule. The dynamic Hebbian learning rule ensures furthermore, that the resulting synaptic strength is dynamically stable, and that interactions between synapses do not result in local instabilities. The rule clearly demonstrates that synapses function as independent localized computational entities, each contributing to the global activity, not in a simply linear fashion, but in a manner that is appropriate to achieve local and global stability of the neuron and the entire dendritic structure.

H-Channels Affect Frequency, Power and Amplitude Fluctuations of Neuronal Network Oscillations.

Oscillations in network activity are ubiquitous in the brain and are involved in diverse cognitive functions. Oscillation characteristics, such as power, frequency, and temporal structure, depend on both network connectivity and intrinsic cellular properties, such as ion channel composition. An important class of channels, with key roles in regulating cell excitability, are h-channels. The h-current (Ih) is a slow, hyperpolarization-activated, depolarizing current that contributes to neuronal resonance and membrane potential. The impact of Ih on network oscillations, however, remains poorly understood. To elucidate the network effects of Ih, we used a computational model of a generic oscillatory neuronal network consisting of inhibitory and excitatory cells that were externally driven by excitatory action potentials and sustained depolarizing currents. We found that Ih increased the oscillation frequency and, in combination with external action potentials, representing input from areas outside the network, strongly decreased the synchrony of firing. As a consequence, the oscillation power and the duration of episodes during which the network exhibited high-amplitude oscillations were greatly reduced in the presence of Ih. Our results suggest that modulation of Ih or impaired expression of h-channels, as observed in epilepsy, could, by affecting oscillation dynamics, markedly alter network-level activity and potentially influence oscillation-dependent cognitive processes such as learning, memory and attention.

Inter-network interactions: impact of connections between oscillatory neuronal networks on oscillation frequency and pattern.

Oscillations in electrical activity are a characteristic feature of many brain networks and display a wide variety of temporal patterns. A network may express a single oscillation frequency, alternate between two or more distinct frequencies, or continually express multiple frequencies. In addition, oscillation amplitude may fluctuate over time. The origin of this complex repertoire of activity remains unclear. Different cortical layers often produce distinct oscillation frequencies. To investigate whether interactions between different networks could contribute to the variety of oscillation patterns, we created two model networks, one generating on its own a relatively slow frequency (20 Hz; slow network) and one generating a fast frequency (32 Hz; fast network). Taking either the slow or the fast network as source network projecting connections to the other, or target, network, we systematically investigated how type and strength of inter-network connections affected target network activity. For high inter-network connection strengths, we found that the slow network was more effective at completely imposing its rhythm on the fast network than the other way around. The strongest entrainment occurred when excitatory cells of the slow network projected to excitatory or inhibitory cells of the fast network. The fast network most strongly imposed its rhythm on the slow network when its excitatory cells projected to excitatory cells of the slow network. Interestingly, for lower inter-network connection strengths, multiple frequencies coexisted in the target network. Just as observed in rat prefrontal cortex, the target network could express multiple frequencies at the same time, alternate between two distinct oscillation frequencies, or express a single frequency with alternating episodes of high and low amplitude. Together, our results suggest that input from other oscillating networks may markedly alter a network's frequency spectrum and may partly be responsible for the rich repertoire of temporal oscillation patterns observed in the brain.

Axonal and dendritic density field estimation from incomplete single-slice neuronal reconstructions.

Neuronal information processing in cortical networks critically depends on the organization of synaptic connectivity. Synaptic connections can form when axons and dendrites come in close proximity of each other. The spatial innervation of neuronal arborizations can be described by their axonal and dendritic density fields. Recently we showed that potential locations of synapses between neurons can be estimated from their overlapping axonal and dendritic density fields. However, deriving density fields from single-slice neuronal reconstructions is hampered by incompleteness because of cut branches. Here, we describe a method for recovering the lost axonal and dendritic mass. This so-called completion method is based on an estimation of the mass inside the slice and an extrapolation to the space outside the slice, assuming axial symmetry in the mass distribution. We validated the method using a set of neurons generated with our NETMORPH simulator. The model-generated neurons were artificially sliced and subsequently recovered by the completion method. Depending on slice thickness and arbor extent, branches that have lost their outside parents (orphan branches) may occur inside the slice. Not connected anymore to the contiguous structure of the sliced neuron, orphan branches result in an underestimation of neurite mass. For 300 µm thick slices, however, the validation showed a full recovery of dendritic and an almost full recovery of axonal mass. The completion method was applied to three experimental data sets of reconstructed rat cortical L2/3 pyramidal neurons. The results showed that in 300 µm thick slices intracortical axons lost about 50% and dendrites about 16% of their mass. The completion method can be applied to single-slice reconstructions as long as axial symmetry can be assumed in the mass distribution. This opens up the possibility of using incomplete neuronal reconstructions from open-access data bases to determine population mean mass density fields.

A morpho-density approach to estimating neural connectivity.

Neuronal signal integration and information processing in cortical neuronal networks critically depend on the organization of synaptic connectivity. Because of the challenges involved in measuring a large number of neurons, synaptic connectivity is difficult to determine experimentally. Current computational methods for estimating connectivity typically rely on the juxtaposition of experimentally available neurons and applying mathematical techniques to compute estimates of neural connectivity. However, since the number of available neurons is very limited, these connectivity estimates may be subject to large uncertainties. We use a morpho-density field approach applied to a vast ensemble of model-generated neurons. A morpho-density field (MDF) describes the distribution of neural mass in the space around the neural soma. The estimated axonal and dendritic MDFs are derived from 100,000 model neurons that are generated by a stochastic phenomenological model of neurite outgrowth. These MDFs are then used to estimate the connectivity between pairs of neurons as a function of their inter-soma displacement. Compared with other density-field methods, our approach to estimating synaptic connectivity uses fewer restricting assumptions and produces connectivity estimates with a lower standard deviation. An important requirement is that the model-generated neurons reflect accurately the morphology and variation in morphology of the experimental neurons used for optimizing the model parameters. As such, the method remains subject to the uncertainties caused by the limited number of neurons in the experimental data set and by the quality of the model and the assumptions used in creating the MDFs and in calculating estimating connectivity. In summary, MDFs are a powerful tool for visualizing the spatial distribution of axonal and dendritic densities, for estimating the number of potential synapses between neurons with low standard deviation, and for obtaining a greater understanding of the relationship between neural morphology and network connectivity.

Competitive dynamics during resource-driven neurite outgrowth.

Neurons form networks by growing out neurites that synaptically connect to other neurons. During this process, neurites develop complex branched trees. Interestingly, the outgrowth of neurite branches is often accompanied by the simultaneous withdrawal of other branches belonging to the same tree. This apparent competitive outgrowth between branches of the same neuron is relevant for the formation of synaptic connectivity, but the underlying mechanisms are unknown. An essential component of neurites is the cytoskeleton of microtubules, long polymers of tubulin dimers running throughout the entire neurite. To investigate whether competition between neurites can emerge from the dynamics of a resource such as tubulin, we developed a multi-compartmental model of neurite growth. In the model, tubulin is produced in the soma and transported by diffusion and active transport to the growth cones at the tip of the neurites, where it is assembled into microtubules to elongate the neurite. Just as in experimental studies, we find that the outgrowth of a neurite branch can lead to the simultaneous retraction of its neighboring branches. We show that these competitive interactions occur in simple neurite morphologies as well as in complex neurite arborizations and that in developing neurons competition for a growth resource such as tubulin can account for the differential outgrowth of neurite branches. The model predicts that competition between neurite branches decreases with path distance between growth cones, increases with path distance from growth cone to soma, and decreases with a higher rate of active transport. Together, our results suggest that competition between outgrowing neurites can already emerge from relatively simple and basic dynamics of a growth resource. Our findings point to the need to test the model predictions and to determine, by monitoring tubulin concentrations in outgrowing neurons, whether tubulin is the resource for which neurites compete.

Independently outgrowing neurons and geometry-based synapse formation produce networks with realistic synaptic connectivity.

Neuronal signal integration and information processing in cortical networks critically depend on the organization of synaptic connectivity. During development, neurons can form synaptic connections when their axonal and dendritic arborizations come within close proximity of each other. Although many signaling cues are thought to be involved in guiding neuronal extensions, the extent to which accidental appositions between axons and dendrites can already account for synaptic connectivity remains unclear. To investigate this, we generated a local network of cortical L2/3 neurons that grew out independently of each other and that were not guided by any extracellular cues. Synapses were formed when axonal and dendritic branches came by chance within a threshold distance of each other. Despite the absence of guidance cues, we found that the emerging synaptic connectivity showed a good agreement with available experimental data on spatial locations of synapses on dendrites and axons, number of synapses by which neurons are connected, connection probability between neurons, distance between connected neurons, and pattern of synaptic connectivity. The connectivity pattern had a small-world topology but was not scale free. Together, our results suggest that baseline synaptic connectivity in local cortical circuits may largely result from accidentally overlapping axonal and dendritic branches of independently outgrowing neurons.

Estimating neuronal connectivity from axonal and dendritic density fields.

Neurons innervate space by extending axonal and dendritic arborizations. When axons and dendrites come in close proximity of each other, synapses between neurons can be formed. Neurons vary greatly in their morphologies and synaptic connections with other neurons. The size and shape of the arborizations determine the way neurons innervate space. A neuron may therefore be characterized by the spatial distribution of its axonal and dendritic "mass." A population mean "mass" density field of a particular neuron type can be obtained by averaging over the individual variations in neuron geometries. Connectivity in terms of candidate synaptic contacts between neurons can be determined directly on the basis of their arborizations but also indirectly on the basis of their density fields. To decide when a candidate synapse can be formed, we previously developed a criterion defining that axonal and dendritic line pieces should cross in 3D and have an orthogonal distance less than a threshold value. In this paper, we developed new methodology for applying this criterion to density fields. We show that estimates of the number of contacts between neuron pairs calculated from their density fields are fully consistent with the number of contacts calculated from the actual arborizations. However, the estimation of the connection probability and the expected number of contacts per connection cannot be calculated directly from density fields, because density fields do not carry anymore the correlative structure in the spatial distribution of synaptic contacts. Alternatively, these two connectivity measures can be estimated from the expected number of contacts by using empirical mapping functions. The neurons used for the validation studies were generated by our neuron simulator NETMORPH. An example is given of the estimation of average connectivity and Euclidean pre- and postsynaptic distance distributions in a network of neurons represented by their population mean density fields.

External drive to inhibitory cells induces alternating episodes of high- and low-amplitude oscillations.

Electrical oscillations in neuronal network activity are ubiquitous in the brain and have been associated with cognition and behavior. Intriguingly, the amplitude of ongoing oscillations, such as measured in EEG recordings, fluctuates irregularly, with episodes of high amplitude alternating with episodes of low amplitude. Despite the widespread occurrence of amplitude fluctuations in many frequency bands and brain regions, the mechanisms by which they are generated are poorly understood. Here, we show that irregular transitions between sub-second episodes of high- and low-amplitude oscillations in the alpha/beta frequency band occur in a generic neuronal network model consisting of interconnected inhibitory and excitatory cells that are externally driven by sustained cholinergic input and trains of action potentials that activate excitatory synapses. In the model, we identify the action potential drive onto inhibitory cells, which represents input from other brain areas and is shown to desynchronize network activity, to be crucial for the emergence of amplitude fluctuations. We show that the duration distributions of high-amplitude episodes in the model match those observed in rat prefrontal cortex for oscillations induced by the cholinergic agonist carbachol. Furthermore, the mean duration of high-amplitude episodes varies in a bell-shaped manner with carbachol concentration, just as in mouse hippocampus. Our results suggest that amplitude fluctuations are a general property of oscillatory neuronal networks that can arise through background input from areas external to the network.

The Flatness of Bifurcations in 3D Dendritic Trees: An Optimal Design.

The geometry of natural branching systems generally reflects functional optimization. A common property is that their bifurcations are planar and that daughter segments do not turn back in the direction of the parent segment. The present study investigates whether this also applies to bifurcations in 3D dendritic arborizations. This question was earlier addressed in a first study of flatness of 3D dendritic bifurcations by Uylings and Smit (1975), who used the apex angle of the right circular cone as flatness measure. The present study was inspired by recent renewed interest in this measure. Because we encountered ourselves shortcomings of this cone angle measure, the search for an optimal measure for flatness of 3D bifurcation was the second aim of our study. Therefore, a number of measures has been developed in order to quantify flatness and orientation properties of spatial bifurcations. All these measures have been expressed mathematically in terms of the three bifurcation angles between the three pairs of segments in the bifurcation. The flatness measures have been applied and evaluated to bifurcations in rat cortical pyramidal cell basal and apical dendritic trees, and to random spatial bifurcations. Dendritic and random bifurcations show significant different flatness measure distributions, supporting the conclusion that dendritic bifurcations are significantly more flat than random bifurcations. Basal dendritic bifurcations also show the property that their parent segments are generally aligned oppositely to the bisector of the angle between their daughter segments, resulting in "symmetrical" configurations. Such geometries may arise when during neuronal development the segments at a newly formed bifurcation are subjected to elastic tensions, which force the bifurcation into an equilibrium planar shape. Apical bifurcations, however, have parent segments oppositely aligned with one of the daughter segments. These geometries arise in the case of side branching from an existing apical main stem. The aligned "apical" parent and "apical" daughter segment form together with the side branch daughter segment already geometrically a flat configuration. These properties are clearly reflected in the flatness measure distributions. Comparison of the different flatness measures made clear that they all capture flatness properties in a different way. Selection of the most appropriate measure thus depends on the question of research. For our purpose of quantifying flatness and orientation of the segments, the dihedral angle ß was found to be the most discriminative and applicable single measure. Alternatively, the parent elevation and azimuth angle formed an orthogonal pair of measures most clearly demonstrating the dendritic bifurcation "symmetry" properties.

An algorithm for finding candidate synaptic sites in computer generated networks of neurons with realistic morphologies.

Neurons make synaptic connections at locations where axons and dendrites are sufficiently close in space. Typically the required proximity is based on the dimensions of dendritic spines and axonal boutons. Based on this principle one can search those locations in networks formed by reconstructed neurons or computer generated neurons. Candidate synapses are then located where axons and dendrites are within a given criterion distance from each other. Both experimentally reconstructed and model generated neurons are usually represented morphologically by piecewise-linear structures (line pieces or cylinders). Proximity tests are then performed on all pairs of line pieces from both axonal and dendritic branches. Applying just a test on the distance between line pieces may result in local clusters of synaptic sites when more than one pair of nearby line pieces from axonal and dendritic branches is sufficient close, and may introduce a dependency on the length scale of the individual line pieces. The present paper describes a new algorithm for defining locations of candidate synapses which is based on the crossing requirement of a line piece pair, while the length of the orthogonal distance between the line pieces is subjected to the distance criterion for testing 3D proximity.

NETMORPH: a framework for the stochastic generation of large scale neuronal networks with realistic neuron morphologies.

We present a simulation framework, called NETMORPH, for the developmental generation of 3D large-scale neuronal networks with realistic neuron morphologies. In NETMORPH, neuronal morphogenesis is simulated from the perspective of the individual growth cone. For each growth cone in a growing axonal or dendritic tree, its actions of elongation, branching and turning are described in a stochastic, phenomenological manner. In this way, neurons with realistic axonal and dendritic morphologies, including neurite curvature, can be generated. Synapses are formed as neurons grow out and axonal and dendritic branches come in close proximity of each other. NETMORPH is a flexible tool that can be applied to a wide variety of research questions regarding morphology and connectivity. Research applications include studying the complex relationship between neuronal morphology and global patterns of synaptic connectivity. Possible future developments of NETMORPH are discussed.

Physiological consequences of selective suppression of synaptic transmission in developing cerebral cortical networks in vitro: differential effects on intrinsically generated bioelectric discharges in a living 'model' system for slow-wave sleep activity.

Within the context of an updated thorough review of the literature concerning activity-dependent cerebro-cortical development, a survey is made of recent experiments which utilize spontaneous spike-trains as the dependent variable in rodent neocortex cultures when synaptic transmission is interfered with during early ontogeny. Emphasis is placed on the complexity of homeostatic adaptations to reduced as well as intensified firing. Two kinds of adaptation are distinguished: (i) rapid recovery (within several hours) towards baseline levels despite sustained blockade of excitatory synaptic transmission, and (ii) the generation of essentially normal firing patterns in cultures assayed in control medium following development in the presence of excitatory receptor blockers. The former category of homeostatic responses is strongly dependent on the type of preparation, with isolated organotypic explants showing greatly limited plasticity in comparison with co-cultures of matching contralateral pieces of cortical tissue. In such co-cultures, compensatory excitatory drive manifests itself even when all three known types of ionotropic glutamate receptors are chronically blocked, and is then mediated by (muscarinic) cholinergic mechanisms which normally do not contribute measurably to spontaneous activity. The rapid return of high levels of spontaneous firing during sustained selective glutamatergic receptor blockade appears to protect neuronal cultures treated in this way from becoming hyperexcitable. In particular, quasi-epileptiform paroxysmal bursting upon return to control medium, such as appears in preparations where bioelectric activity has been totally suppressed during network formation, fails to appear in chronically receptor blocked cultures. On the contrary, desensitization of blocked glutamate receptors, as a physiological compensation for the up-regulation of non-blocked receptors, could be demonstrated for both the AMPA and the NMDA glutamate receptor sub-types. This wide range of homeostatic responses underscores the importance of spontaneous neuronal discharges for setting and maintaining an optimal balance between excitatory and inhibitory mechanisms in developing neocortical networks.

Low-frequency stimulation induces stable transitions in stereotypical activity in cortical networks.

Reverberating spontaneous synchronized brain activity is believed to play an important role in neural information processing. Whether and how external stimuli can influence this spontaneous activity is poorly understood. Because periodic synchronized network activity is also prominent in in vitro neuronal cultures, we used cortical cultures grown on multielectrode arrays to examine how spontaneous activity is affected by external stimuli. Spontaneous network activity before and after low-frequency electrical stimulation was quantified in several ways. Our results show that the initially stable pattern of stereotypical spontaneous activity was transformed into another activity pattern that remained stable for at least 1 h. The transformations consisted of changes in single site and culture-wide network activity as well as in the spatiotemporal dynamics of network bursting. We show for the first time that low-frequency electrical stimulation can induce long-lasting alterations in spontaneous activity of cortical neuronal networks. We discuss whether the observed transformations in network activity could represent a switch in attractor state.

Latency dependent development of related firing patterns of cultured cortical neurons.

Networks of cortical neurons were grown over multi electrode arrays to enable simultaneous measurement of signals from multiple neurons. We described functional connectivity in these networks by relationships between individual electrodes, based on conditional firing probabilities. In this study we investigated periods in which the strength of a relationship monotonously increased (strengthening) or decreased (weakening) during periods of 24 or 10 hours. We observed a slightly increased incidence of latencies up to 25 ms during strengthening, while these latencies rarely occurred during weakening. Next, it appeared that relationships tended to strengthen more than average in periods with latencies in the range 5-30 ms, whereas strengthening was significantly less than average in latencies 40-65 ms.

Homeostatically regulated spontaneous neuronal discharges protect developing cerebral cortex networks from becoming hyperactive following prolonged blockade of excitatory synaptic receptors.

In order to further examine the role of spontaneous action potential (SAP) discharges in neocortical development, amino-acid-mediated synaptic transmission was selectively blocked in an improved organotypic neocortex culture preparation. Contralateral occipital cortex slices from neonatal rats were co-cultured for several weeks in a ventricle-to-ventricle orientation known to greatly enhance cyto-morphological and electrophysiological maturation. Such preparations are highly resistant to attempts to suppress neuronal firing by blocking ionotropic glutamate receptors: not only can kainate receptors partly substitute for NMDA- and AMPA-mediated neurotransmission when these receptors are pharmacologically blocked, but (muscarinic) cholinergic receptors also begin to drive SAP activity when the kainate receptors, too, are chronically blocked. Only tetrodotoxin proved able to eliminate SAPs altogether in these co-cultures, while GABAergic receptor blockade (using bicucculine) led to persistent epileptiform discharges. Treatment effects were assayed upon transfer to control medium by means of a quantitative analysis of spontaneously occurring polyneuronal spike trains. Total suppression of action potentials for several weeks (by tetrodotoxin treatment) led, as in earlier experiments, to strongly intensified burst firing upon transfer to control medium. Chronic glutamate receptor blocked cultures, on the other hand, showed only minor deviations from control firing levels and patterns when assayed in normal medium. Protection against the development of hyperactivity despite partial blockade of synaptic transmission was roughly proportional to the degree to which spontaneous firing during the treatment period approximated normal SAP levels. This homeostatic response to chronically reduced excitatory drive thus differs from earlier results using isolated organotypic cortex cultures, in which the restoration of SAP activity failed to prevent the development of network hyperactivity. Chronic bicucculine treatment, in contrast, had little or no homeostatic effect on SAP firing patterns; on the contrary, opposite to earlier findings using isolated occipital cortex explants, paroxysmal discharges persisted even after transfer to control medium.

Spontaneous neuronal discharge patterns in developing organotypic mega-co-cultures of neonatal rat cerebral cortex.

Sagittal slices of neonatal rat neocortex, extending from the prefrontal to the occipital area, were cultured separately or in pairs, oriented in such a way that axons projecting from the ventricular surface of each explant could innervate the other one. Functional connections were made between as well as within the explants, and spontaneous field potentials and associated action potentials occurred in variable bursts, and with varying degrees of synchrony. Spike-train analysis revealed that the activity patterns seen in these 'mega' co-cultures closely mimic 'tracé alternant' patterns, consisting of trains of burst discharges recurring several times per minute, which are characteristic for the immature intact cerebral cortex during slow-wave sleep. The prefrontal region was consistently less active than the occipital area but the two were qualitatively similar with respect to their patterns of neuronal firing. Isolated mega-cultures, on the other hand, despite their large size, exhibited only intermittent brief bursts that closely resembled those observed both in occipital cortex tissue fragments and in dissociated cell cultures. The mega-co-culture preparation thus appears to give the best currently available approximation to intrinsic cerebral discharge patterns in vivo.

Compartment volume influences microtubule dynamic instability: a model study.

Microtubules (MTs) are cytoskeletal polymers that exhibit dynamic instability, the random alternation between growth and shrinkage. MT dynamic instability plays an essential role in cell development, division, and motility. To investigate dynamic instability, simulation models have been widely used. However, conditions under which the concentration of free tubulin fluctuates as a result of growing or shrinking MTs have not been studied before. Such conditions can arise, for example, in small compartments, such as neuronal growth cones. Here we investigate by means of computational modeling how concentration fluctuations caused by growing and shrinking MTs affect dynamic instability. We show that these fluctuations shorten MT growth and shrinkage times and change their distributions from exponential to non-exponential, gamma-like. Gamma-like distributions of MT growth and shrinkage times, which allow optimal stochastic searching by MTs, have been observed in various cell types and are believed to require structural changes in the MT during growth or shrinkage. Our results, however, show that these distributions can already arise as a result of fluctuations in the concentration of free tubulin due to growing and shrinking MTs. Such fluctuations are possible not only in small compartments but also when tubulin diffusion is slow or when many MTs (de)polymerize synchronously. Volume and all other factors that influence these fluctuations can affect MT dynamic instability and, consequently, the processes that depend on it, such as neuronal growth cone behavior and cell motility in general.

Dynamics and plasticity in developing neuronal networks in vitro.

When dissociated cortical tissue is brought into culture, neurons readily grow out by forming axonal and dendritic arborizations and synaptic connections. These developing neuronal networks in vitro display spontaneous firing activity from about the end of the first week in vitro. When cultured on multielectrode arrays firing activity can be recorded from many neurons simultaneously over long periods of time. These experimental approaches provide valuable data for studying firing dynamics in neuronal networks in relation to an ongoing development of neurons and synaptic connectivity in the network. This chapter summarizes recent findings on the characteristics and developmental changes in the spontaneous firing dynamics. These changes include long-lasting transient periods of increased firing at individual sites on a time scale of days to weeks, and an age-specific repetitive pattern of synchronous network firing (network bursts) on a time scale of seconds. Especially the spatio-temporal organization of firing within network bursts showed great stability over many hours. In addition, a progressive day-to-day evolution was observed, with an initial broadening of the burst firing rate profile during the 3rd week in vitro (WIV) and a pattern of abrupt onset and precise spike timing from the 5th WIV onwards. These developmental changes are discussed in the light of structural changes in the network and activity-dependent plasticity mechanisms. Preliminary findings are presented on the pattern of spike sequences within network burst, as well as the effect of external stimulation on the spatio-temporal organization within network bursts.

Compensatory physiological responses to chronic blockade of amino acid receptors during early development in spontaneously active organotypic cerebral cortex explants cultured in vitro.

Paired organotypic explants from rat occipital cortex were cultured for up to three weeks in the presence of selective blockers of amino acid receptor blockers, during which period spontaneous action potential generation was monitored electrophysiologically. In contrast to isolated explants (Corner, M.A., van Pelt, J., Wolters, P.S., Baker, R.E.and Nuytinck, R.H. (2002) Physiological e.ects of sustained blockade of excitatory synaptic transmission on spontaneously active developing neuronal networks--an inquiry into the reciprocal linkage between intrinsic biorhythms and neuroplasticity in early ontogeny. Neurosci. Biobehav. Rev., 26: 127-185), which upregulated their initially depressed spontaneous bursting activity only under conditions of N-methyl D-aspartate (NMDA) receptor blockade, cross-innervated co-cultures showed a large degree of functional recovery even when combined NMDA and AMPA receptor blockade was carried out. This compensatory activity could be eliminated by acute addition of a selective kainate receptor blocker to the medium. When kainate along with AMPA and NMDA receptor mediated activity was chronically suppressed, however, considerable functional recovery--in the form of recurrent burst discharges--took place gradually over a period of three weeks in vitro. These spontaneous bursts disappeared rapidly upon treatment with the muscarinic receptor blocker, atropine, but continuous low-level firing emerged at the same time. Similar "tonic" background activity was induced in control cultures as well, but without any noticeable reduction in burst discharges. Co-cultured neocortex explants, in which cyto-morphological maturation proceeds to a far greater degree than in isolated explants (Baker, R.E.and van Pelt, J. (1997) Co-cultured but not isolated cortical explants display normal dendritic development: a longterm quantitative study. Dev. Brain Res., 98: 21-27) are evidently capable of an astonishing degree of functional compensation for loss of excitatory synaptic drive during development. It could be shown, furthermore, that such homeostatic responses are not mediated largely by a weakening of inhibitory mechanisms in the absence of spontaneous firing. Chronic inhibitory synaptic blockade, on the other hand, led to intensified bursting activity which gradually normalized over a 3-week culture period. The cellular basis for this reversal of the disinhibited state, as well as for the residual neuronal firing even after cholinergic mechanisms have been largely eliminated, is at present unknown. The degree to which immature cortical networks attempt to compensate for altered levels of physiological activity, as documented in the present report, is another indication of how important such activity can be for normal development (see Corner, M.A., van Pelt, J., Wolters, P.S., Baker, R.E. and Nuytinck, R.H. (2002) Physiological e.ects of sustained blockade of excitatory synaptic transmission on spontaneously active developing neuronal networks-an inquiry into the reciprocal linkage between intrinsic biorhythms and neuroplasticity in early ontogeny. Neurosci. Biobehav. Rev., 26: 127-185).. At the same time, the large variations in overall firing levels and "macro-scale" temporal patterns from culture to culture within a given series, despite all attempts at identical preparation of the explants, can only mean that the "set-points" for such regulation are themselves subject to unknown ontogenetic factors which, apparently, are nonuniformly distributed even within a restricted region of the neocortex. On the other hand, it was striking to note that, regardless of age or treatment, an unexpected degree of consistency in temporal patterning existed at "mini-" and "micro-" time-scales (viz., EEG delta and beta frequency ranges, respectively) even when network bursting tendencies became greatly reduced in favor of tonic firing.