Temporal modulation of spike-timing-dependent plasticity

Fast learning without synaptic plasticity in spiking neural networks

Spiking neural networks are of high current interest, both from the perspective of modelling neural networks of the brain and for porting their fast learning capability and energy efficiency into neuromorphic hardware. But so far we have not been able to reproduce fast learning capabilities of the brain in spiking neural networks. Biological data suggest that a synergy of synaptic plasticity on a slow time scale with network dynamics on a faster time scale is responsible for fast learning capabilities of the brain. We show here that a suitable orchestration of this synergy between synaptic plasticity and network dynamics does in fact reproduce fast learning capabilities of generic recurrent networks of spiking neurons. This points to the important role of recurrent connections in spiking networks, since these are necessary for enabling salient network dynamics. We show more specifically that the proposed synergy enables synaptic weights to encode more general information such as priors and task structures, since moment-to-moment processing of new information can be delegated to the network dynamics.

Sequence anticipation and spike-timing-dependent plasticity emerge from a predictive learning rule

Intelligent behavior depends on the brain's ability to anticipate future events. However, the learning rules that enable neurons to predict and fire ahead of sensory inputs remain largely unknown. We propose a plasticity rule based on predictive processing, where the neuron learns a low-rank model of the synaptic input dynamics in its membrane potential. Neurons thereby amplify those synapses that maximally predict other synaptic inputs based on their temporal relations, which provide a solution to an optimization problem that can be implemented at the single-neuron level using only local information. Consequently, neurons learn sequences over long timescales and shift their spikes towards the first inputs in a sequence. We show that this mechanism can explain the development of anticipatory signalling and recall in a recurrent network. Furthermore, we demonstrate that the learning rule gives rise to several experimentally observed STDP (spike-timing-dependent plasticity) mechanisms. These findings suggest prediction as a guiding principle to orchestrate learning and synaptic plasticity in single neurons.

Brain is modulated by neuronal plasticity during postnatal development

Neuroplasticity is referred to the ability of the nervous system to change its structure or functions as a result of former stimuli. It is a plausible mechanism underlying a dynamic brain through adaptation processes of neural structure and activity patterns. Nevertheless, it is still unclear how the plastic neural systems achieve and maintain their equilibrium. Additionally, the alterations of balanced brain dynamics under different plasticity rules have not been explored either. Therefore, the present article primarily aims to review recent research studies regarding homosynaptic and heterosynaptic neuroplasticity characterized by the manipulation of excitatory and inhibitory synaptic inputs. Moreover, it attempts to understand different mechanisms related to the main forms of synaptic plasticity at the excitatory and inhibitory synapses during the brain development processes. Hence, this study comprised surveying those articles published since 1988 and available through PubMed, Google Scholar and science direct databases on a keyword-based search paradigm. All in all, the study results presented extensive and corroborative pieces of evidence for the main types of plasticity, including the long-term potentiation (LTP) and long-term depression (LTD) of the excitatory and inhibitory postsynaptic potentials (EPSPs and IPSPs).

Heterosynaptic Plasticity Determines the Set Point for Cortical Excitatory-Inhibitory Balance

Excitation in neural circuits must be carefully controlled by inhibition to regulate information processing and network excitability. During development, cortical inhibitory and excitatory inputs are initially mismatched but become co-tuned or balanced with experience. However, little is known about how excitatory-inhibitory balance is defined at most synapses or about the mechanisms for establishing or maintaining this balance at specific set points. Here we show how coordinated long-term plasticity calibrates populations of excitatory-inhibitory inputs onto mouse auditory cortical pyramidal neurons. Pairing pre- and postsynaptic activity induced plasticity at paired inputs and different forms of heterosynaptic plasticity at the strongest unpaired synapses, which required minutes of activity and dendritic Ca²⁺ signaling to be computed. Theoretical analyses demonstrated how the relative rate of heterosynaptic plasticity could normalize and stabilize synaptic strengths to achieve any possible excitatory-inhibitory correlation. Thus, excitatory-inhibitory balance is dynamic and cell specific, determined by distinct plasticity rules across multiple excitatory and inhibitory synapses.

Visual Contrast Modulates Operant Learning Responses in Larval Zebrafish

The larval zebrafish is a promising vertebrate model organism to study neural mechanisms underlying learning and memory due to its small brain and rich behavioral repertoire. Here, we report on a high-throughput operant conditioning system for zebrafish larvae, which can simultaneously train 12 fish to associate a visual conditioned pattern with electroshocks. We find that the learning responses can be enhanced by the visual contrast, not the spatial features of the conditioned patterns, highlighted by several behavioral metrics. By further characterizing the learning curves as well as memory extinction, we demonstrate that the percentage of learners and the memory length increase as the conditioned pattern becomes darker. Finally, little difference in operant learning responses was found between AB wild-type fish and elavl3:H2B-GCaMP6f transgenic fish.

Familiarity Detection is an Intrinsic Property of Cortical Microcircuits with Bidirectional Synaptic Plasticity

Humans instantly recognize a previously seen face as “familiar.” To deepen our understanding of familiarity-novelty detection, we simulated biologically plausible neural network models of generic cortical microcircuits consisting of spiking neurons with random recurrent synaptic connections. NMDA receptor (NMDAR)-dependent synaptic plasticity was implemented to allow for unsupervised learning and bidirectional modifications. Network spiking activity evoked by sensory inputs consisting of face images altered synaptic efficacy, which resulted in the network responding more strongly to a previously seen face than a novel face. Network size determined how many faces could be accurately recognized as familiar. When the simulated model became sufficiently complex in structure, multiple familiarity traces could be retained in the same network by forming partially-overlapping subnetworks that differ slightly from each other, thereby resulting in a high storage capacity. Fisher’s discriminant analysis was applied to identify critical neurons whose spiking activity predicted familiar input patterns. Intriguingly, as sensory exposure was prolonged, the selected critical neurons tended to appear at deeper layers of the network model, suggesting recruitment of additional circuits in the network for incremental information storage. We conclude that generic cortical microcircuits with bidirectional synaptic plasticity have an intrinsic ability to detect familiar inputs. This ability does not require a specialized wiring diagram or supervision and can therefore be expected to emerge naturally in developing cortical circuits.

Demonstrating Hybrid Learning in a Flexible Neuromorphic Hardware System

We present results from a new approach to learning and plasticity in neuromorphic hardware systems: to enable flexibility in implementable learning mechanisms while keeping high efficiency associated with neuromorphic implementations, we combine a general-purpose processor with full-custom analog elements. This processor is operating in parallel with a fully parallel neuromorphic system consisting of an array of synapses connected to analog, continuous time neuron circuits. Novel analog correlation sensor circuits process spike events for each synapse in parallel and in real-time. The processor uses this pre-processing to compute new weights possibly using additional information following its program. Therefore, to a certain extent, learning rules can be defined in software giving a large degree of flexibility. Synapses realize correlation detection geared towards Spike-Timing Dependent Plasticity (STDP) as central computational primitive in the analog domain. Operating at a speed-up factor of 1000 compared to biological time-scale, we measure time-constants from tens to hundreds of micro-seconds. We analyze variability across multiple chips and demonstrate learning using a multiplicative STDP rule. We conclude that the presented approach will enable flexible and efficient learning as a platform for neuroscientific research and technological applications.

Modulation of Synaptic Plasticity by Glutamatergic Gliotransmission: A Modeling Study

Glutamatergic gliotransmission, that is, the release of glutamate from perisynaptic astrocyte processes in an activity-dependent manner, has emerged as a potentially crucial signaling pathway for regulation of synaptic plasticity, yet its modes of expression and function in vivo remain unclear. Here, we focus on two experimentally well-identified gliotransmitter pathways, (i) modulations of synaptic release and (ii) postsynaptic slow inward currents mediated by glutamate released from astrocytes, and investigate their possible functional relevance on synaptic plasticity in a biophysical model of an astrocyte-regulated synapse. Our model predicts that both pathways could profoundly affect both short- and long-term plasticity. In particular, activity-dependent glutamate release from astrocytes could dramatically change spike-timing-dependent plasticity, turning potentiation into depression (and vice versa) for the same induction protocol.

Brain-controlled neuromuscular stimulation to drive neural plasticity and functional recovery

There is mounting evidence that appropriately timed neuromuscular stimulation can induce neural plasticity and generate functional recovery from motor disorders. This review addresses the idea that coordinating stimulation with a patient's voluntary effort might further enhance neurorehabilitation. Studies in cell cultures and behaving animals have delineated the rules underlying neural plasticity when single neurons are used as triggers. However, the rules governing more complex stimuli and larger networks are less well understood. We argue that functional recovery might be optimized if stimulation were modulated by a brain machine interface, to match the details of the patient's voluntary intent. The potential of this novel approach highlights the need for a better understanding of the complex rules underlying this form of plasticity.

Plasticity of cortical excitatory-inhibitory balance

Synapses are highly plastic and are modified by changes in patterns of neural activity or sensory experience. Plasticity of cortical excitatory synapses is thought to be important for learning and memory, leading to alterations in sensory representations and cognitive maps. However, these changes must be coordinated across other synapses within local circuits to preserve neural coding schemes and the organization of excitatory and inhibitory inputs, i.e., excitatory-inhibitory balance. Recent studies indicate that inhibitory synapses are also plastic and are controlled directly by a large number of neuromodulators, particularly during episodes of learning. Many modulators transiently alter excitatory-inhibitory balance by decreasing inhibition, and thus disinhibition has emerged as a major mechanism by which neuromodulation might enable long-term synaptic modifications naturally. This review examines the relationships between neuromodulation and synaptic plasticity, focusing on the induction of long-term changes that collectively enhance cortical excitatory-inhibitory balance for improving perception and behavior.

Inhibitory and excitatory spike-timing-dependent plasticity in the auditory cortex

Synapses are plastic and can be modified by changes in spike timing. Whereas most studies of long-term synaptic plasticity focus on excitation, inhibitory plasticity may be critical for controlling information processing, memory storage, and overall excitability in neural circuits. Here we examine spike-timing-dependent plasticity (STDP) of inhibitory synapses onto layer 5 neurons in slices of mouse auditory cortex, together with concomitant STDP of excitatory synapses. Pairing pre- and postsynaptic spikes potentiated inhibitory inputs irrespective of precise temporal order within ∼10 ms. This was in contrast to excitatory inputs, which displayed an asymmetrical STDP time window. These combined synaptic modifications both required NMDA receptor activation and adjusted the excitatory-inhibitory ratio of events paired with postsynaptic spiking. Finally, subthreshold events became suprathreshold, and the time window between excitation and inhibition became more precise. These findings demonstrate that cortical inhibitory plasticity requires interactions with co-activated excitatory synapses to properly regulate excitatory-inhibitory balance.

Dopamine D1 and D5 receptors modulate spike timing-dependent plasticity at medial perforant path to dentate granule cell synapses

Although evidence suggests that DA modulates hippocampal function, the mechanisms underlying that dopaminergic modulation are largely unknown. Using perforated-patch electrophysiological techniques to maintain the intracellular milieu, we investigated how the activation of D1-type DA receptors regulates spike timing-dependent plasticity (STDP) of the medial perforant path (mPP) synapse onto dentate granule cells. When D1-type receptors were inhibited, a relatively mild STDP protocol induced LTP only within a very narrow timing window between presynaptic stimulation and postsynaptic response. The stimulus protocol produced timing-dependent LTP (tLTP) only when the presynaptic stimulation was followed 30 ms later by depolarization-induced postsynaptic action potentials. That is, the time between presynaptic stimulation and postsynaptic response was 30 ms (Δt = +30 ms). When D1-type receptors were activated, however, the same mild STDP protocol induced tLTP over a much broader timing window: tLTP was induced when -30 ms ≤ Δt ≤ +30 ms. The result indicated that D1-type receptor activation enabled synaptic potentiation even when postsynaptic activity preceded presynaptic stimulation within this Δt range. Results with null mice lacking the Kv4.2 potassium channel and with the potassium channel inhibitor, 4-aminopyridine, suggested that D1-type receptors enhanced tLTP induction by suppressing the transient IA-type K(+) current. Results obtained with antagonists and DA receptor knock-out mice indicated that endogenous activity of both D1 and D5 receptors modulated plasticity in the mPP. The DA D5 receptors appeared particularly important in regulating plasticity of the mPP onto the dentate granule cells.

Volitional muscle activity paired with transcranial magnetic stimulation increases corticospinal excitability

Studies of activity-dependent stimulation in non-human primates suggest that pairing each instance of volitional muscle activity with immediate intracortical stimulation causes long-term-potentiation-like effects. This technique holds promise for clinical rehabilitation, yet few investigators have tested activity-dependent stimulation in human subjects. In addition, no one has studied activity-dependent stimulation on the cortical representation for two separate target muscles in human subjects. We hypothesized that 40 min of transcranial magnetic stimulation (TMS) triggered from ballistic muscle activity at a mean repetition rate of 1 Hz would cause greater increases in corticospinal excitability than TMS-cued muscle activity, and that these changes would be specific to the muscle of study. Ten healthy human subjects participated in 4 separate sessions in this crossover study: (1) visually cued volitional activation of the abductor pollicis brevis (APB) muscle triggering TMS (APB-Triggered TMS), (2) volitional activation of APB in response to TMS delivered from a recording of the prior APB-Triggered TMS session (TMS-Cued APB), (3) visually cued volitional activation of the extensor digitorum (ED) triggering TMS (ED-Triggered TMS), and (4) volitional activation of ED in response to TMS delivered from a recording of the prior ED-Triggered TMS session (TMS-Cued ED). Contrary to our hypothesis, we discovered evidence of increased corticospinal excitability for all conditions as measured by change in area of the motor evoked potential. We conclude that single TMS pulses paired either before or after muscle activity may increase corticospinal excitability and that further studies are needed to clarify the optimal time window for inducing neural plasticity with activity-dependent stimulation. These findings will inform the design of future activity-dependent stimulation protocols for clinical rehabilitation.

Activity-dependent synaptic plasticity of a chalcogenide electronic synapse for neuromorphic systems

Nanoscale inorganic electronic synapses or synaptic devices, which are capable of emulating the functions of biological synapses of brain neuronal systems, are regarded as the basic building blocks for beyond-Von Neumann computing architecture, combining information storage and processing. Here, we demonstrate a Ag/AgInSbTe/Ag structure for chalcogenide memristor-based electronic synapses. The memristive characteristics with reproducible gradual resistance tuning are utilised to mimic the activity-dependent synaptic plasticity that serves as the basis of memory and learning. Bidirectional long-term Hebbian plasticity modulation is implemented by the coactivity of pre- and postsynaptic spikes, and the sign and degree are affected by assorted factors including the temporal difference, spike rate and voltage. Moreover, synaptic saturation is observed to be an adjustment of Hebbian rules to stabilise the growth of synaptic weights. Our results may contribute to the development of highly functional plastic electronic synapses and the further construction of next-generation parallel neuromorphic computing architecture.

Biologically inspired load balancing mechanism in neocortical competitive learning

A unique delayed self-inhibitory pathway mediated by layer 5 Martinotti Cells was studied in a biologically inspired neural network simulation. Inclusion of this pathway along with layer 5 basket cell lateral inhibition caused balanced competitive learning, which led to the formation of neuronal clusters as were indeed reported in the same region. Martinotti pathway proves to act as a learning “conscience,” causing overly successful regions in the network to restrict themselves and let others fire. It thus spreads connectivity more evenly throughout the net and solves the “dead unit” problem of clustering algorithms in a local and biologically plausible manner.

Spike-timing-dependent plasticity in primate corticospinal connections induced during free behavior

Motor learning and functional recovery from brain damage involve changes in the strength of synaptic connections between neurons. Relevant in vivo evidence on the underlying cellular mechanisms remains limited and indirect. We found that the strength of neural connections between motor cortex and spinal cord in monkeys can be modified with an autonomous recurrent neural interface that delivers electrical stimuli in the spinal cord triggered by action potentials of corticospinal cells during free behavior. The activity-dependent stimulation modified the strength of the terminal connections of single corticomotoneuronal cells, consistent with a bidirectional spike-timing-dependent plasticity rule previously derived from in vitro experiments. For some cells, the changes lasted for days after the end of conditioning, but most effects eventually reverted to preconditioning levels. These results provide direct evidence of corticospinal synaptic plasticity in vivo at the level of single neurons induced by normal firing patterns during free behavior.

Adrenergic gating of Hebbian spike-timing-dependent plasticity in cortical interneurons

In pyramidal cells, the induction of spike-dependent plasticity (STDP) follows a simple Hebbian rule in which the order of presynaptic and postsynaptic firing dictates the induction of LTP or LTD. In contrast, cortical fast spiking (FS) interneurons, which control the rate and timing of pyramidal cell firing, reportedly express timing-dependent LTD, but not timing-dependent LTP. Because a mismatch in STDP rules could impact the maintenance of the excitation/inhibition balance, we examined the neuromodulation of STDP in FS cells of mouse visual cortex. We found that stimulation of adrenergic receptors enables the induction of Hebbian bidirectional STDP in FS cells in a manner consistent with a pull-push mechanism previously characterized in pyramidal cells. However, in pyramidal cells, STDP induction depends on NMDA receptors, whereas in FS cells it depends on mGluR5 receptors. We propose that neuromodulators control the polarity of STDP in different synapses in the same manner, and independently of the induction mechanism, by acting downstream in the plasticity cascade. By doing so, neuromodulators may allow coordinated plastic changes in FS and pyramidal cells.

The Convallis rule for unsupervised learning in cortical networks

The phenomenology and cellular mechanisms of cortical synaptic plasticity are becoming known in increasing detail, but the computational principles by which cortical plasticity enables the development of sensory representations are unclear. Here we describe a framework for cortical synaptic plasticity termed the "Convallis rule", mathematically derived from a principle of unsupervised learning via constrained optimization. Implementation of the rule caused a recurrent cortex-like network of simulated spiking neurons to develop rate representations of real-world speech stimuli, enabling classification by a downstream linear decoder. Applied to spike patterns used in in vitro plasticity experiments, the rule reproduced multiple results including and beyond STDP. However STDP alone produced poorer learning performance. The mathematical form of the rule is consistent with a dual coincidence detector mechanism that has been suggested by experiments in several synaptic classes of juvenile neocortex. Based on this confluence of normative, phenomenological, and mechanistic evidence, we suggest that the rule may approximate a fundamental computational principle of the neocortex.

Reward-based learning under hardware constraints-using a RISC processor embedded in a neuromorphic substrate

In this study, we propose and analyze in simulations a new, highly flexible method of implementing synaptic plasticity in a wafer-scale, accelerated neuromorphic hardware system. The study focuses on globally modulated STDP, as a special use-case of this method. Flexibility is achieved by embedding a general-purpose processor dedicated to plasticity into the wafer. To evaluate the suitability of the proposed system, we use a reward modulated STDP rule in a spike train learning task. A single layer of neurons is trained to fire at specific points in time with only the reward as feedback. This model is simulated to measure its performance, i.e., the increase in received reward after learning. Using this performance as baseline, we then simulate the model with various constraints imposed by the proposed implementation and compare the performance. The simulated constraints include discretized synaptic weights, a restricted interface between analog synapses and embedded processor, and mismatch of analog circuits. We find that probabilistic updates can increase the performance of low-resolution weights, a simple interface between analog synapses and processor is sufficient for learning, and performance is insensitive to mismatch. Further, we consider communication latency between wafer and the conventional control computer system that is simulating the environment. This latency increases the delay, with which the reward is sent to the embedded processor. Because of the time continuous operation of the analog synapses, delay can cause a deviation of the updates as compared to the not delayed situation. We find that for highly accelerated systems latency has to be kept to a minimum. This study demonstrates the suitability of the proposed implementation to emulate the selected reward modulated STDP learning rule. It is therefore an ideal candidate for implementation in an upgraded version of the wafer-scale system developed within the BrainScaleS project.

Topological dynamics in spike-timing dependent plastic model neural networks

Spike-timing dependent plasticity (STDP) is a biologically constrained unsupervised form of learning that potentiates or depresses synaptic connections based on the precise timing of pre-synaptic and post-synaptic firings. The effects of on-going STDP on the topology of evolving model neural networks were assessed in 50 unique simulations which modeled 2 h of activity. After a period of stabilization, a number of global and local topological features were monitored periodically to quantify on-going changes in network structure. Global topological features included the total number of remaining synapses, average synaptic strengths, and average number of synapses per neuron (degree). Under a range of different input regimes and initial network configurations, each network maintained a robust and highly stable global structure across time. Local topology was monitored by assessing state changes of all three-neuron subgraphs (triads) present in the networks. Overall counts and the range of triad configurations varied little across the simulations; however, a substantial set of individual triads continued to undergo rapid state changes and revealed a dynamic local topology. In addition, specific small-world properties also fluctuated across time. These findings suggest that on-going STDP provides an efficient means of selecting and maintaining a stable yet flexible network organization.

Dark exposure extends the integration window for spike-timing-dependent plasticity

Metaplasticity, the adaptive changes of long-term potentiation (LTP) and long-term depression (LTD) in response to fluctuations in neural activity is well documented in visual cortex, where dark rearing shifts the frequency threshold for the induction of LTP and LTD. Here we studied metaplasticity affecting spike-timing-dependent plasticity, in which the polarity of plasticity is determined not by the stimulation frequency, but by the temporal relationship between near-coincidental presynaptic and postsynaptic firing. We found that in mouse visual cortex the same regime of deprivation that restricts the frequency range for inducing rate-dependent LTD extends the integration window for inducing timing-dependent LTD, enabling LTD induction with random presynaptic and postsynaptic firing. Notably, the underlying mechanism for the changes in both rate-dependent and time-dependent LTD appears to be an increase of NR2b-containing NMDAR at the synapse. Thus, the rules of metaplasticity might manifest in opposite directions, depending on the plasticity-induction paradigms.

Bursts modify electrical synaptic strength

Changes in synaptic strength resulting from neuronal activity have been described in great detail for chemical synapses, but the relationship between natural forms of activity and the strength of electrical synapses had previously not been investigated. The thalamic reticular nucleus (TRN), a brain area rich in gap junctional (electrical) synapses, regulates cortical attention, initiates sleep spindles, and participates in shifts between states of arousal. Plasticity of electrical synapses in the TRN may be a key mechanism underlying these processes. Recently, we demonstrated a novel activity-dependent form of long-term depression of electrical synapses in the TRN (Haas et al., 2011). Here we provide an overview of those findings and discuss them in broader context. Because gap junctional proteins are widely expressed in the mammalian brain, modification of synaptic strength is likely to be a widespread and powerful mechanism at electrical synapses throughout the brain.

The spike-timing dependence of plasticity

In spike-timing-dependent plasticity (STDP), the order and precise temporal interval between presynaptic and postsynaptic spikes determine the sign and magnitude of long-term potentiation (LTP) or depression (LTD). STDP is widely utilized in models of circuit-level plasticity, development, and learning. However, spike timing is just one of several factors (including firing rate, synaptic cooperativity, and depolarization) that govern plasticity induction, and its relative importance varies across synapses and activity regimes. This review summarizes this broader view of plasticity, including the forms and cellular mechanisms for the spike-timing dependence of plasticity, and, the evidence that spike timing is an important determinant of plasticity in vivo.

The pharmacology of neuroplasticity induced by non-invasive brain stimulation: building models for the clinical use of CNS active drugs

The term neuroplasticity encompasses structural and functional modifications of neuronal connectivity. Abnormal neuroplasticity is involved in various neuropsychiatric diseases, such as dystonia, epilepsy, migraine, Alzheimer's disease, fronto-temporal degeneration, schizophrenia, and post cerebral stroke. Drugs affecting neuroplasticity are increasingly used as therapeutics in these conditions. Neuroplasticity was first discovered and explored in animal experimentation. However, non-invasive brain stimulation (NIBS) has enabled researchers recently to induce and study similar processes in the intact human brain. Plasticity induced by NIBS can be modulated by pharmacological interventions, targeting ion channels, or neurotransmitters. Importantly, abnormalities of plasticity as studied by NIBS are directly related to clinical symptoms in neuropsychiatric diseases. Therefore, a core theme of this review is the hypothesis that NIBS-induced plasticity can explore and potentially predict the therapeutic efficacy of CNS-acting drugs in neuropsychiatric diseases. We will (a) review the basics of neuroplasticity, as explored in animal experimentation, and relate these to our knowledge about neuroplasticity induced in humans by NIBS techniques. We will then (b) discuss pharmacological modulation of plasticity in animals and humans. Finally, we will (c) review abnormalities of plasticity in neuropsychiatric diseases, and discuss how the combination of NIBS with pharmacological intervention may improve our understanding of the pathophysiology of abnormal plasticity in these diseases and their purposeful pharmacological treatment.

Learning complex temporal patterns with resource-dependent spike timing-dependent plasticity

Studies of spike timing-dependent plasticity (STDP) have revealed that long-term changes in the strength of a synapse may be modulated substantially by temporal relationships between multiple presynaptic and postsynaptic spikes. Whereas long-term potentiation (LTP) and long-term depression (LTD) of synaptic strength have been modeled as distinct or separate functional mechanisms, here, we propose a new shared resource model. A functional consequence of our model is fast, stable, and diverse unsupervised learning of temporal multispike patterns with a biologically consistent spiking neural network. Due to interdependencies between LTP and LTD, dendritic delays, and proactive homeostatic aspects of the model, neurons are equipped to learn to decode temporally coded information within spike bursts. Moreover, neurons learn spike timing with few exposures in substantial noise and jitter. Surprisingly, despite having only one parameter, the model also accurately predicts in vitro observations of STDP in more complex multispike trains, as well as rate-dependent effects. We discuss candidate commonalities in natural long-term plasticity mechanisms.

Optophysiological approach to resolve neuronal action potentials with high spatial and temporal resolution in cultured neurons

Cell to cell communication in the central nervous system is encoded into transient and local membrane potential changes (ΔVm). Deciphering the rules that govern synaptic transmission and plasticity entails to be able to perform V_m recordings throughout the entire neuronal arborization. Classical electrophysiology is, in most cases, not able to do so within small and fragile neuronal subcompartments. Thus, optical techniques based on the use of fluorescent voltage-sensitive dyes (VSDs) have been developed. However, reporting spontaneous or small ΔV_m from neuronal ramifications has been challenging, in part due to the limited sensitivity and phototoxicity of VSD-based optical measurements. Here we demonstrate the use of water soluble VSD, ANNINE-6plus, with laser-scanning microscopy to optically record ΔV_m in cultured neurons. We show that the sensitivity (>10% of fluorescence change for 100 mV depolarization) and time response (sub millisecond) of the dye allows the robust detection of action potentials (APs) even without averaging, allowing the measurement of spontaneous neuronal firing patterns. In addition, we show that back-propagating APs can be recorded, along distinct dendritic sites and within dendritic spines. Importantly, our approach does not induce any detectable phototoxic effect on cultured neurons. This optophysiological approach provides a simple, minimally invasive, and versatile optical method to measure electrical activity in cultured neurons with high temporal (ms) resolution and high spatial (μm) resolution.

Stability versus neuronal specialization for STDP: long-tail weight distributions solve the dilemma

Spike-timing-dependent plasticity (STDP) modifies the weight (or strength) of synaptic connections between neurons and is considered to be crucial for generating network structure. It has been observed in physiology that, in addition to spike timing, the weight update also depends on the current value of the weight. The functional implications of this feature are still largely unclear. Additive STDP gives rise to strong competition among synapses, but due to the absence of weight dependence, it requires hard boundaries to secure the stability of weight dynamics. Multiplicative STDP with linear weight dependence for depression ensures stability, but it lacks sufficiently strong competition required to obtain a clear synaptic specialization. A solution to this stability-versus-function dilemma can be found with an intermediate parametrization between additive and multiplicative STDP. Here we propose a novel solution to the dilemma, named log-STDP, whose key feature is a sublinear weight dependence for depression. Due to its specific weight dependence, this new model can produce significantly broad weight distributions with no hard upper bound, similar to those recently observed in experiments. Log-STDP induces graded competition between synapses, such that synapses receiving stronger input correlations are pushed further in the tail of (very) large weights. Strong weights are functionally important to enhance the neuronal response to synchronous spike volleys. Depending on the input configuration, multiple groups of correlated synaptic inputs exhibit either winner-share-all or winner-take-all behavior. When the configuration of input correlations changes, individual synapses quickly and robustly readapt to represent the new configuration. We also demonstrate the advantages of log-STDP for generating a stable structure of strong weights in a recurrently connected network. These properties of log-STDP are compared with those of previous models. Through long-tail weight distributions, log-STDP achieves both stable dynamics for and robust competition of synapses, which are crucial for spike-based information processing.

A history of spike-timing-dependent plasticity

How learning and memory is achieved in the brain is a central question in neuroscience. Key to today's research into information storage in the brain is the concept of synaptic plasticity, a notion that has been heavily influenced by Hebb's (1949) postulate. Hebb conjectured that repeatedly and persistently co-active cells should increase connective strength among populations of interconnected neurons as a means of storing a memory trace, also known as an engram. Hebb certainly was not the first to make such a conjecture, as we show in this history. Nevertheless, literally thousands of studies into the classical frequency-dependent paradigm of cellular learning rules were directly inspired by the Hebbian postulate. But in more recent years, a novel concept in cellular learning has emerged, where temporal order instead of frequency is emphasized. This new learning paradigm - known as spike-timing-dependent plasticity (STDP) - has rapidly gained tremendous interest, perhaps because of its combination of elegant simplicity, biological plausibility, and computational power. But what are the roots of today's STDP concept? Here, we discuss several centuries of diverse thinking, beginning with philosophers such as Aristotle, Locke, and Ribot, traversing, e.g., Lugaro's plasticità and Rosenblatt's perceptron, and culminating with the discovery of STDP. We highlight interactions between theoretical and experimental fields, showing how discoveries sometimes occurred in parallel, seemingly without much knowledge of the other field, and sometimes via concrete back-and-forth communication. We point out where the future directions may lie, which includes interneuron STDP, the functional impact of STDP, its mechanisms and its neuromodulatory regulation, and the linking of STDP to the developmental formation and continuous plasticity of neuronal networks.

Timing is not Everything: Neuromodulation Opens the STDP Gate

Spike timing dependent plasticity (STDP) is a temporally specific extension of Hebbian associative plasticity that has tied together the timing of presynaptic inputs relative to the postsynaptic single spike. However, it is difficult to translate this mechanism to in vivo conditions where there is an abundance of presynaptic activity constantly impinging upon the dendritic tree as well as ongoing postsynaptic spiking activity that backpropagates along the dendrite. Theoretical studies have proposed that, in addition to this pre- and postsynaptic activity, a “third factor” would enable the association of specific inputs to specific outputs. Experimentally, the picture that is beginning to emerge, is that in addition to the precise timing of pre- and postsynaptic spikes, this third factor involves neuromodulators that have a distinctive influence on STDP rules. Specifically, neuromodulatory systems can influence STDP rules by acting via dopaminergic, noradrenergic, muscarinic, and nicotinic receptors. Neuromodulator actions can enable STDP induction or – by increasing or decreasing the threshold – can change the conditions for plasticity induction. Because some of the neuromodulators are also involved in reward, a link between STDP and reward-mediated learning is emerging. However, many outstanding questions concerning the relationship between neuromodulatory systems and STDP rules remain, that once solved, will help make the crucial link from timing-based synaptic plasticity rules to behaviorally based learning.

Questions about STDP as a General Model of Synaptic Plasticity

According to spike-timing-dependent plasticity (STDP), the timing of the Na⁺ spike relative to the EPSP determines whether LTP or LTD will occur. Here, we review our reservations about STDP. Most investigations of this process have been done under conditions in which the spike is evoked by postsynaptic current injection. Under more realistic conditions, in which the spike is evoked by the EPSP, the results do not generally support STDP. For instance, low-frequency stimulation of a group of synapses can cause LTD, not the LTP predicted by the pre-before-post sequence in STDP; this is true regardless of whether or not the EPSP is large enough to produce a Na⁺ spike. With stronger or more frequent stimulation, LTP can be induced by the same pre-before-post timing, but in this case block of Na⁺ spikes does not necessarily prevent LTP induction. Thus, Na⁺ spikes may facilitate LTP and/or LTD under some conditions, but they are not necessary, a finding consistent with their small size relative to the EPSP in many parts of pyramidal cell dendrites. The nature of the dendritic depolarizing events that control bidirectional plasticity is of central importance to understanding neural function. There are several candidates, including backpropagating action potentials, but also dendritic Ca²⁺ spikes, the AMPA receptor-mediated EPSP, and NMDA receptor-mediated EPSPs or spikes. These often appear to be more important than the Na⁺ spike in providing the depolarization necessary for plasticity. We thus feel that it is premature to accept STDP-like processes as the major determinant of LTP/LTD.

Dendritic synapse location and neocortical spike-timing-dependent plasticity

While it has been appreciated for decades that synapse location in the dendritic tree has a powerful influence on signal processing in neurons, the role of dendritic synapse location on the induction of long-term synaptic plasticity has only recently been explored. Here, we review recent work revealing how learning rules for spike-timing-dependent plasticity (STDP) in cortical neurons vary with the spatial location of synaptic input. A common principle appears to be that proximal synapses show conventional STDP, whereas distal inputs undergo plasticity according to novel learning rules. One crucial factor determining location-dependent STDP is the backpropagating action potential, which tends to decrease in amplitude and increase in width as it propagates into the dendritic tree of cortical neurons. We discuss additional location-dependent mechanisms as well as the functional implications of heterogeneous learning rules at different dendritic locations for the organization of synaptic inputs.