Heterosynaptic Competition Research Articles

Role of primary afferents in the developmental regulation of motor axon synapse numbers on Renshaw cells.

Motor function in mammalian species depends on the maturation of spinal circuits formed by a large variety of interneurons that regulate motoneuron firing and motor output. Interneuron activity is in turn modulated by the organization of their synaptic inputs, but the principles governing the development of specific synaptic architectures unique to each premotor interneuron are unknown. For example, Renshaw cells receive, at least in the neonate, convergent inputs from sensory afferents (likely Ia) and motor axons, raising the question of whether they interact during Renshaw cell development. In other well-studied neurons, such as Purkinje cells, heterosynaptic competition between inputs from different sources shapes synaptic organization. To examine the possibility that sensory afferents modulate synaptic maturation on developing Renshaw cells, we used three animal models in which afferent inputs in the ventral horn are dramatically reduced (ER81(-/-) knockout), weakened (Egr3(-/-) knockout), or strengthened (mlcNT3(+/-) transgenic). We demonstrate that increasing the strength of sensory inputs on Renshaw cells prevents their deselection and reduces motor axon synaptic density, and, in contrast, absent or diminished sensory afferent inputs correlate with increased densities of motor axons synapses. No effects were observed on other glutamatergic inputs. We conclude that the early strength of Ia synapses influences their maintenance or weakening during later development and that heterosynaptic influences from sensory synapses during early development regulates the density and organization of motor inputs on mature Renshaw cells.

From Blickets to Synapses: Inferring Temporal Causal Networks by Observation

How do human infants learn the causal dependencies between events? Evidence suggests that this remarkable feat can be achieved by observation of only a handful of examples. Many computational models have been produced to explain how infants perform causal inference without explicit teaching about statistics or the scientific method. Here, we propose a spiking neuronal network implementation that can be entrained to form a dynamical model of the temporal and causal relationships between events that it observes. The network uses spike-time dependent plasticity, long-term depression, and heterosynaptic competition rules to implement Rescorla-Wagner-like learning. Transmission delays between neurons allow the network to learn a forward model of the temporal relationships between events. Within this framework, biologically realistic synaptic plasticity rules account for well-known behavioral data regarding cognitive causal assumptions such as backwards blocking and screening-off. These models can then be run as emulators for state inference. Furthermore, this mechanism is capable of copying synaptic connectivity patterns between neuronal networks by observing the spontaneous spike activity from the neuronal circuit that is to be copied, and it thereby provides a powerful method for transmission of circuit functionality between brain regions.

A Hebbian learning rule gives rise to mirror neurons and links them to control theoretic inverse models

Mirror neurons are neurons whose responses to the observation of a motor act resemble responses measured during production of that act. Computationally, mirror neurons have been viewed as evidence for the existence of internal inverse models. Such models, rooted within control theory, map-desired sensory targets onto the motor commands required to generate those targets. To jointly explore both the formation of mirrored responses and their functional contribution to inverse models, we develop a correlation-based theory of interactions between a sensory and a motor area. We show that a simple eligibility-weighted Hebbian learning rule, operating within a sensorimotor loop during motor explorations and stabilized by heterosynaptic competition, naturally gives rise to mirror neurons as well as control theoretic inverse models encoded in the synaptic weights from sensory to motor neurons. Crucially, we find that the correlational structure or stereotypy of the neural code underlying motor explorations determines the nature of the learned inverse model: random motor codes lead to causal inverses that map sensory activity patterns to their motor causes; such inverses are maximally useful, by allowing the imitation of arbitrary sensory target sequences. By contrast, stereotyped motor codes lead to less useful predictive inverses that map sensory activity to future motor actions. Our theory generalizes previous work on inverse models by showing that such models can be learned in a simple Hebbian framework without the need for error signals or backpropagation, and it makes new conceptual connections between the causal nature of inverse models, the statistical structure of motor variability, and the time-lag between sensory and motor responses of mirror neurons. Applied to bird song learning, our theory can account for puzzling aspects of the song system, including necessity of sensorimotor gating and selectivity of auditory responses to bird's own song (BOS) stimuli.

Shared Synaptic Pathophysiology in Syndromic and Nonsyndromic Rodent Models of Autism

The genetic heterogeneity of autism poses a major challenge for identifying mechanism-based treatments. A number of rare mutations are associated with autism, and it is unclear whether these result in common neuronal alterations. Monogenic syndromes, such as fragile X, include autism as one of their multifaceted symptoms and have revealed specific defects in synaptic plasticity. We discovered an unexpected convergence of synaptic pathophysiology in a nonsyndromic form of autism with those in fragile X syndrome. Neuroligin-3 knockout mice (a model for nonsyndromic autism) exhibited disrupted heterosynaptic competition and perturbed metabotropic glutamate receptor-dependent synaptic plasticity, a hallmark of fragile X. These phenotypes could be rescued by reexpression of neuroligin-3 in juvenile mice, highlighting the possibility of reverting neuronal circuit alterations in autism after the completion of development.

Cav2.1 in Cerebellar Purkinje Cells Regulates Competitive Excitatory Synaptic Wiring, Cell Survival, and Cerebellar Biochemical Compartmentalization

In the adult cerebellum, each Purkinje cell (PC) is innervated by a single climbing fiber (CF) in proximal dendrites and 10(5)-10(6) parallel fibers (PFs) in distal dendrites. This organized wiring is established postnatally through heterosynaptic competition between PFs and CFs and homosynaptic competition among multiple CFs. Using PC-specific Cav2.1 knock-out mice (PC-Cav2.1 KO mice), we have demonstrated recently that postsynaptic Cav2.1 plays a key role in the homosynaptic competition by promoting functional strengthening and dendritic translocation of single "winner" CFs. Here, we report that Cav2.1 in PCs, but not in granule cells, is also essential for the heterosynaptic competition. In PC-Cav2.1 KO mice, the extent of CF territory was limited to the soma and basal dendrites, whereas PF territory was expanded reciprocally. Consequently, the proximal somatodendritic domain of PCs displayed hyperspiny transformation and fell into chaotic innervation by multiple CFs and numerous PFs. PC-Cav2.1 KO mice also displayed patterned degeneration of PCs, which occurred preferentially in aldolase C/zebrin II-negative cerebellar compartments. Furthermore, the mutually complementary expression of phospholipase Cβ3 (PLCβ3) and PLCβ4 was altered such that their normally sharp boundary was blurred in the PCs of PC-Cav2.1 KO mice. This blurring was caused by an impaired posttranscriptional downregulation of PLCβ3 in PLCβ4-dominant PCs during the early postnatal period. A similar alteration was noted in the banded expression of the glutamate transporter EAAT4 in PC-Cav2.1 KO mice. Therefore, Cav2.1 in PCs is essential for competitive synaptic wiring, cell survival, and the establishment of precise boundaries and reciprocity of biochemical compartments in PCs.

Development of an anatomical technique for visualizing the mode of climbing fiber innervation in Purkinje cells and its application to mutant mice lacking GluRδ2 and Cav2.1

In the adult cerebellum, a single climbing fiber (CF) innervates proximal dendrites of Purkinje cells (PCs). This monoinnervation is established by the developmental elimination of surplus CFs through homosynaptic competition among multiply innervating CFs and heterosynaptic competition between CFs and parallel fibers, i.e., granule cell axons innervating distal PC dendrites. Although the developmental process of CF monoinnervation and defects in it in mutant and experimental animal models have been extensively studied by electrophysiological techniques, for quite some time this subject was poorly understood from a morphological perspective due to a lack of neuroanatomical methods that could distinguish CFs with different neuronal origins. Soon after the identification of type 2 vesicular glutamate transporter (VGluT2) that selectively detects CF terminals in the molecular layer, we developed a novel method of combined anterograde tracer labeling and VGluT2 immunohistochemistry. This method enables us to identify the mode (mono vs. multiple) of CF innervation and the site of multiple innervation. Since then, we have applied this method to various kinds of gene-manipulated mice manifesting ataxia and other cerebellar phenotypes. In this review, we summarize experimental procedures for the combined tracer/VGluT2 labeling method, and then introduce what we have learned by applying this method in studies on the role of GluRδ2 and Ca(v)2.1 in CF monoinnervation. This method has provided informative anatomical correlates to electrophysiological data and vice versa, and will extend our knowledge of the molecular and cellular mechanisms for the development, plasticity, degeneration, and repair of the CF-PC projection system.

Ablation of Glutamate Receptor GluRδ2 in Adult Purkinje Cells Causes Multiple Innervation of Climbing Fibers by Inducing Aberrant Invasion to Parallel Fiber Innervation Territory

Glutamate receptor GluRδ2 is exclusively expressed in Purkinje cells (PCs) from early development and plays key roles in parallel fiber (PF) synapse formation, elimination of surplus climbing fibers (CFs), long-term depression, motor coordination, and motor learning. To address its role in adulthood, we previously developed a mouse model of drug-induced GluRδ2 ablation in adult PCs (Takeuchi et al., 2005). In that study, we demonstrated an essential role to maintain the connectivity of PF-PC synapses, based on the observation that both mismatching of presynaptic and postsynaptic specializations and disconnection of PF-PC synapses are progressively increased after GluRδ2 ablation. Here, we pursued its role for CF wiring in adult cerebellum. In parallel with the disconnection of PF-PC synapses, ascending CF branches exhibited distal extension to innervate distal dendrites of the target and neighboring PCs. Furthermore, transverse CF branches, a short motile collateral rarely forming synapses in wild-type animals, displayed aberrant mediolateral extension to innervate distal dendrites of neighboring and remote PCs. Consequently, many PCs were wired by single main CF and other surplus CFs innervating a small part of distal dendrites. Electrophysiological recording further revealed that surplus CF-EPSCs characterized with slow rise time and small amplitude emerged after GluRδ2 ablation, and increased progressively both in number and amplitude. Therefore, GluRδ2 is essential for maintaining CF monoinnervation in adult cerebellum by suppressing aberrant invasion of CF branches to the territory of PF innervation. Thus, GluRδ2 fuels heterosynaptic competition and gives PFs the competitive advantages over CFs throughout the animal's life.

Spike-Time-Dependent Plasticity and Heterosynaptic Competition Organize Networks to Produce Long Scale-Free Sequences of Neural Activity

Sequential neural activity patterns are as ubiquitous as the outputs they drive, which include motor gestures and sequential cognitive processes. Neural sequences are long, compared to the activation durations of participating neurons, and sequence coding is sparse. Numerous studies demonstrate that spike-time-dependent plasticity (STDP), the primary known mechanism for temporal order learning in neurons, cannot organize networks to generate long sequences, raising the question of how such networks are formed. We show that heterosynaptic competition within single neurons, when combined with STDP, organizes networks to generate long unary activity sequences even without sequential training inputs. The network produces a diversity of sequences with a power law length distribution and exponent -1, independent of cellular time constants. We show evidence for a similar distribution of sequence lengths in the recorded premotor song activity of songbirds. These results suggest that neural sequences may be shaped by synaptic constraints and network circuitry rather than cellular time constants.

GluRdelta2 expression in the mature cerebellum of hotfoot mice promotes parallel fiber synaptogenesis and axonal competition.

Glutamate receptor delta 2 (GluRδ2) is selectively expressed in the cerebellum, exclusively in the spines of the Purkinje cells (PCs) that are in contact with parallel fibers (PFs). Although its structure is similar to ionotropic glutamate receptors, it has no channel function and its ligand is unknown. The GluRδ2-null mice, such as knockout and hotfoot have profoundly altered cerebellar circuitry, which causes ataxia and impaired motor learning. Notably, GluRδ2 in PC-PF synapses regulates their maturation and strengthening and induces long term depression (LTD). In addition, GluRδ2 participates in the highly territorial competition between the two excitatory inputs to the PC; the climbing fiber (CF), which innervates the proximal dendritic compartment, and the PF, which is connected to spiny distal branchlets. Recently, studies have suggested that GluRδ2 acts as an adhesion molecule in PF synaptogenesis. Here, we provide in vivo and in vitro evidence that supports this hypothesis. Through lentiviral rescue in hotfoot mice, we noted a recovery of PC-PF contacts in the distal dendritic domain. In the proximal domain, we observed the formation of new spines that were innervated by PFs and a reduction in contact with the CF; ie, the pattern of innervation in the PC shifted to favor the PF input. Moreover, ectopic expression of GluRδ2 in HEK293 cells that were cocultured with granule cells or in cerebellar Golgi cells in the mature brain induced the formation of new PF contacts. Collectively, our observations show that GluRδ2 is an adhesion molecule that induces the formation of PF contacts independently of its cellular localization and promotes heterosynaptic competition in the PC proximal dendritic domain.

Molecular Mechanisms Governing Competitive Synaptic Wiring in Cerebellar Purkinje Cells

Cerebellar Purkinje cells (PCs) play a principal role in motor coordination and motor learning. To fulfill these functions, PCs receive and integrate two types of excitatory inputs, climbing fiber (CF) and parallel fiber (PF). CFs are projection axons from the inferior olive, and convey error signals to PCs. On the other hand, PFs are T-shaped axons of cerebellar granule cells, and convey sensory and motor information carried through the pontocerebellar and spinocerebellar mossy fiber pathways. The most remarkable feature of PC circuits is the highly territorial innervation by these two excitatory afferents. A single climbing CF powerfully and exclusively innervates proximal PC dendrites, whereas hundreds of thousands of PFs innervate distal PC dendrites. Recent studies using gene-manipulated mice have been elucidating that the PC circuitry is formed and maintained by molecular mechanisms that fuel homosynaptic competition among CFs and heterosynaptic competition between CFs and PFs. GluRdelta2 (a PC-specific glutamate receptor) and precerebellin or Cbln1 (a granule cell-derived secretory protein) cooperatively work for selective strengthening of PF-PC synapses, and prevent excessive distal extension of CFs that eventually causes multiple innervation at distal dendrites. In contrast, P/Q-type Ca2+ channels, which mediate Ca2+ influx upon CF activity, selectively strengthen the innervation by a single main CF, and expel PFs and other CFs from proximal dendrites that it innervates. Therefore, we now understand that owing to these mechanisms, territorial innervation by CFs and PFs is properly structured and mono-innervation by CFs is established. Several key issues for future study are also discussed.

Experience-Driven Axon Retraction without Binocular Imbalance in Developing Visual Cortex

Refinement of the neural circuit during brain maturation is regulated by experience-driven neural activity. In the mammalian visual cortex, monocular visual deprivation (MD) in the early postnatal life causes a significant loss of cortical responses to a deprived eye and the retraction of input axons serving the deprived eye. A competitive interaction between inputs serving both eyes has been supposed to underlie the effects of MD because the loss of cortical response is much weaker when both eyes are deprived of vision. Also, the input axons do not retract after binocular deprivation. Here, we report that uncorrelated activity between presynaptic and postsynaptic neurons can solely lead to the retraction of geniculocortical axons in the absence of activity imbalance between two inputs. We analyzed the morphology of geniculocortical axons in a pharmacologically inhibited visual cortex of animals with normal vision and of binocularly deprived animals. In the normal vision animals, the axonal arbors in the inhibited cortex showed robust retraction. On the other hand, the arbors in binocularly deprived animals remained mostly intact. These results suggest that a homosynaptic associative mechanism, rather than a heterosynaptic competition between inputs, may play an important role in experience-driven axon retraction.

P/Q-type Ca2+ channel alpha1A regulates synaptic competition on developing cerebellar Purkinje cells.

Synapse formation depends critically on the competition among inputs of multiple sources to individual neurons. Cerebellar Purkinje cells have highly organized synaptic wiring from two distinct sources of excitatory afferents. Single climbing fibers innervate proximal dendrites of Purkinje cells, whereas numerous parallel fibers converge on their distal dendrites. Here, we demonstrate that the P/Q-type Ca2+ channel alpha1A, a major Ca2+ channel subtype in Purkinje cells, is crucial for this organized synapse formation. In the alpha1A knock-out mouse, many ectopic spines were protruded from proximal dendrites and somata of Purkinje cells. Innervation territory of parallel fibers was expanded proximally to innervate the ectopic spines, whereas that of climbing fibers was regressed to the basal portion of proximal dendrites and somata. Furthermore, multiple climbing fibers consisting of a strong climbing fiber and one or a few weaker climbing fibers, persisted in the majority of Purkinje cells and were cowired to the same somata, proximal dendrites, or both. Therefore, the lack of alpha1A results in the persistence of parallel fibers and surplus climbing fibers, which should normally be expelled from the compartment innervated by the main climbing fiber. These results suggest that a P/Q-type Ca2+ channel alpha1A fuels heterosynaptic competition between climbing fibers and parallel fibers and also fuels homosynaptic competition among multiple climbing fibers. This molecular function facilitates the distal extension of climbing fiber innervation along the dendritic tree of the Purkinje cell and also establishes climbing fiber monoinnervation of individual Purkinje cells.

Stable Neural Attractors Formation: Learning Rules and Network Dynamics

Different models of attractor networks have been proposed to form cell assemblies. Among them, networks with a fixed synaptic matrix can be distinguished from those including learning dynamics, since the latter adapt the attractor landscape of the lateral connections according to the statistics of the presented stimuli, yielding a more complex behavior. We propose a new learning rule that builds internal representations of input timuli as attractors of neurons in a recurrent network. The dynamics of activation and synaptic adaptation are analyzed in experiments where representations for different input patterns are formed, focusing on the properties of the model as a memory system. The experimental results are exposed along with a survey of different Hebbian rules proposed in the literature for attractors formation. These rules are compared with the help of a new tool, the learning map, where LTP and LTD, as well as homo- and heterosynaptic competition, can be graphically interpreted.

Segregation of ON and OFF retinogeniculate connectivity directed by patterned spontaneous activity.

In many parts of the developing nervous system, the early patterns of connectivity are refined by processes that require neuronal activity. These processes are thought to involve Hebbian mechanisms that lead to strengthening and maintenance of inputs that display correlated pre- and postsynaptic activity and elimination of inputs that fire asynchronously. Here we investigated the role of patterned spontaneous retinal activity and Hebbian synaptic mechanisms on segregation of ON and OFF retinal afferents in the dorsal lateral geniculate nucleus (dLGN) of the developing ferret visual system. We recorded extracellularly the spontaneous spike activity of neighboring pairs of ganglion cells and found that OFF cells have significantly higher mean firing rates than ON cells. Spiking is best correlated between cells of the same sign (ON, ON; OFF, OFF) compared with cells of opposite sign (ON, OFF). We then constructed a simple Hebbian model of retinogeniculate synaptic development based on a correlational framework. Using our recorded activity patterns, together with previous calcium-imaging data, we show that endogenous retinal activity, coupled with Hebbian mechanisms of synaptic development, can drive the segregation of ON and OFF retinal inputs to the dLGN. Segregation occurs robustly when heterosynaptic competition is present within time windows of 50-500 ms. In addition, our results suggest that the initial patterns of connectivity (biases in convergence of inputs) and the strength of inhibition in the network each play a crucial role in determining whether ON or OFF inputs dominate at maturity.

Cortical plasticity: Is it time for a change?

Classical studies of plasticity in the visual cortex have been interpreted in terms of heterosynaptic competition between inputs. But an alternative type of ‘homosynaptic’ plasticity can explain many recent observations and has recently received experimental support. Perhaps both types of plasticity are important.

Long-term depression of excitatory synaptic transmission and its relationship to long-term potentiation.

In many brain areas, including the cerebellar cortex, neocortex, hippocampus, striatum and nucleus accumbens, brief activation of an excitatory pathway can produce long-term depression (LTD) of synaptic transmission. In most preparations, induction of LTD has been shown to require a minimum level of postsynaptic depolarization and a rise in the intracellular Ca2+ concentration [Ca2+]i in the postsynaptic neurone. Thus, induction conditions resemble those described for the initiation of associative long-term potentiation (LTP). However, data from structures susceptible to both LTD and LTP suggest that a stronger depolarization and a greater increase in [Ca2+]i are required to induce LTP than to initiate LTD. The source of Ca2+ appears to be less critical for the differential induction of LTP and LTD than the amplitude of the Ca2+ surge, since the activation of voltage- and ligand-gated Ca2+ conductances as well as the release from intracellular stores have all been shown to contribute to both LTD and LTP induction. LTD is induceable even at inactive synapses if [Ca2+]i is raised to the appropriate level by antidromic or heterosynaptic activation, or by raising the extracellular Ca2+ concentration [Ca2+]o. These conditions suggest a rule (called here the ABS rule) for activity-dependent synaptic modifications that differs from the classical Hebb rule and that can account for both homosynaptic LTD and LTP as well as for heterosynaptic competition and associativity.