Spike-timing Dependent Synaptic Plasticity Research Articles

High-Performance Memristive Synapse Composed of Ferroelectric ZnVO-Based Schottky Junction.

In pursuit of realizing neuromorphic computing devices, we demonstrated the high-performance synaptic functions on the top-to-bottom Au/ZnVO/Pt two-terminal ferroelectric Schottky junction (FSJ) device architecture. The active layer of ZnVO exhibited the ferroelectric characteristics because of the broken lattice-translational symmetry, arising from the incorporation of smaller V5+ ions into smaller Zn2+ host lattice sites. The fabricated FSJ devices displayed an asymmetric hysteresis behavior attributed to the ferroelectric polarization-dependent Schottky field-emission rate difference in between positive and negative bias voltage regions. Additionally, it was observed that the magnitude of the on-state current could be systematically controlled by changing either the amplitude or the width of the applied voltage pulses. Owing to these voltage pulse-tunable multi-state memory characteristics, the device revealed diverse synaptic functions such as short-term memory, dynamic range-tunable long-term memory, and versatile rules in spike time-dependent synaptic plasticity. For the pattern-recognition simulation, furthermore, more than 95% accuracy was recorded when using the optimized experimental device parameters. These findings suggest the ZnVO-based FSJ device holds significant promise for application in next-generation brain-inspired neuromorphic computing systems.

Nonlinear Dynamics in HfO2/SiO2-Based Interface Dipole Modulation Field-Effect Transistors for Synaptic Applications

In the pursuit of energy-efficient spiking neural network (SNN) hardware, synaptic devices leveraging emerging memory technologies hold significant promise. This study investigates the application of the recently proposed HfO2/SiO2-based interface dipole modulation (IDM) memory for synaptic spike timing-dependent plasticity (STDP) learning. Firstly, through pulse measurements of IDM metal–oxide–semiconductor (MOS) capacitors, we demonstrate that IDM exhibits an inherently nonlinear and near-symmetric response. Secondly, we discuss the drain current response of a field-effect transistor (FET) incorporating a multi-stack IDM structure, revealing its nonlinear and asymmetric pulse response, and suggest that the degree of the asymmetry depends on the modulation current ratio. Thirdly, to emulate synaptic STDP behavior, we implement double-pulse-controlled drain current modulation of IDMFET using a simple bipolar rectangular pulse. Additionally, we propose a double-pulse-controlled synaptic depression that is valuable for optimizing STDP-based unsupervised learning. Integrating the pulse response characteristics of IDMFETs into a two-layer SNN system for synaptic weight updates, we assess training and classification performance on handwritten digits. Our results demonstrate that IDMFET-based synaptic devices can achieve classification accuracy comparable to previously reported simulation-based results.

GluN2B-NMDAR subunit contribution on synaptic plasticity: A phenomenological model for CA3-CA1 synapses.

Synaptic plasticity is believed to be a key mechanism underlying learning and memory. We developed a phenomenological N-methyl-D-aspartate (NMDA) receptor-based voltage-dependent synaptic plasticity model for synaptic modifications at hippocampal CA3-CA1 synapses on a hippocampal CA1 pyramidal neuron. The model incorporates the GluN2A-NMDA and GluN2B-NMDA receptor subunit-based functions and accounts for the synaptic strength dependence on the postsynaptic NMDA receptor composition and functioning without explicitly modeling the NMDA receptor-mediated intracellular calcium, a local trigger of synaptic plasticity. We embedded the model into a two-compartmental model of a hippocampal CA1 pyramidal cell and validated it against experimental data of spike-timing-dependent synaptic plasticity (STDP), high and low-frequency stimulation. The developed model predicts altered learning rules in synapses formed on the apical dendrites of the detailed compartmental model of CA1 pyramidal neuron in the presence of the GluN2B-NMDA receptor hypofunction and can be used in hippocampal networks to model learning in health and disease.

Altered closed state inactivation gating in Kv4.2 channels results in developmental and epileptic encephalopathies in human patients.

Kv4.2 subunits, encoded by KCND2, serve as the pore-forming components of voltage-gated, inactivating ISA K+ channels expressed in the brain. ISA channels inactivate without opening in response to subthreshold excitatory input, temporarily increasing neuronal excitability, the back propagation of action potentials, and Ca2+ influx into dendrites, thereby regulating mechanisms of spike timing-dependent synaptic plasticity. As previously described, a de novo variant in Kv4.2, p.Val404Met, is associated with an infant-onset developmental and epileptic encephalopathyin monozygotic twin boys. The p.Val404Met variant enhances inactivation directly from closed states, but dramatically impairs inactivation after channel opening. We now report the identification of a closely related, novel, de novo variant in Kv4.2, p.Val402Leu, in a boy with an early-onset pharmacoresistant epilepsy that evolved to an epileptic aphasia syndrome (Continuous Spike Wave during Sleep Syndrome). Like p.Val404Met, the p.Val402Leu variant increases the rate of inactivation from closed states, but significantly slows inactivation after the pore opens. Although quantitatively the p.Val402Leu mutation alters channel kinetics less dramatically than p.Val404Met, our results strongly support the conclusion that p.Val402Leu and p.Val404Met cause the clinical features seen in the affected individuals and underscore the importance of closed state inactivation in ISA channels in normal brain development and function.

Calcium and Spike Timing-Dependent Plasticity.

Since its discovery, spike timing-dependent synaptic plasticity (STDP) has been thought to be a primary mechanism underlying the brain’s ability to learn and to form new memories. However, despite the enormous interest in both the experimental and theoretical neuroscience communities in activity-dependent plasticity, it is still unclear whether plasticity rules inferred from in vitro experiments apply to in vivo conditions. Among the multiple reasons why plasticity rules in vivo might differ significantly from in vitro studies is that extracellular calcium concentration use in most studies is higher than concentrations estimated in vivo. STDP, like many forms of long-term synaptic plasticity, strongly depends on intracellular calcium influx for its induction. Here, we discuss the importance of considering physiological levels of extracellular calcium concentration to study functional plasticity.

Simulation model of CA1 pyramidal neurons reveal opposing roles for the Na+/Ca2+ exchange current and Ca2+-activated K+ current during spike-timing dependent synaptic plasticity.

Sodium Calcium exchanger (NCX) proteins utilize the electrochemical gradient of Na+ to generate Ca2+ efflux (forward mode) or influx (reverse mode). In mammals, there are three unique NCX encoding genes—NCX1, NCX2, and NCX3, that comprise the SLC8A family, and mRNA from all three exchangers is expressed in hippocampal pyramidal cells. Furthermore, mutant ncx2-/- and ncx3-/- mice have each been shown to exhibit altered long-term potentiation (LTP) in the hippocampal CA1 region due to delayed Ca2+ clearance after depolarization that alters synaptic transmission. In addition to the role of NCX at the synapse of hippocampal subfields required for LTP, the three NCX isoforms have also been shown to localize to the dendrite of hippocampal pyramidal cells. In the case of NCX1, it has been shown to localize throughout the basal and apical dendrite of CA1 neurons where it helps compartmentalize Ca2+ between dendritic shafts and spines. Given the role for NCX and calcium in synaptic plasticity, the capacity of NCX splice-forms to influence backpropagating action potentials has clear consequences for the induction of spike-timing dependent synaptic plasticity (STDP). To explore this, we examined the effect of NCX localization, density, and allosteric activation on forward and back propagating signals and, next employed a STDP paradigm to monitor the effect of NCX on plasticity using back propagating action potentials paired with EPSPs. From our simulation studies we identified a role for the sodium calcium exchange current in normalizing STDP, and demonstrate that NCX is required at the postsynaptic site for this response. We also screened other mechanisms in our model and identified a role for the Ca2+ activated K+ current at the postsynapse in producing STDP responses. Together, our data reveal opposing roles for the Na+/Ca2+ exchanger current and the Ca2+ activated K+ current in setting STDP.

Efficient VLSI Architecture for Odor Recognition with a Spiking Neural Network

In this paper spiking neural network (SNN) is presented which can discriminate odor data. Spike timing dependent synaptic plasticity (STDP) means a plasticity which is controlled by the presynaptic and postsynaptic spikes time difference. Using this STDP rule the synaptic weights are modified after the mitral and before the cortical cells. In order to determine whether the circuit has correctly identified the odor the SNN has either a high or a low response at the output for any odor given as the input.

Spike-shape dependence of the spike-timing dependent synaptic plasticity in ferroelectric-tunnel-junction synapses

Resistive switching (RS) devices have attracted increasing attention for artificial synapse applications in neural networks because of their nonvolatile and analogue resistance changes. Among the neural networks, a spiking neural network (SNN) based on spike-timing-dependent plasticity (STDP) is highly energy efficient. To implement STDP in resistive switching devices, several types of voltage spikes have been proposed to date, but there have been few reports on the relationship between the STDP characteristics and spike types. Here, we report the STDP characteristics implemented in ferroelectric tunnel junctions (FTJs) by several types of spikes. Based on simulated time evolutions of superimposed spikes and taking the nonlinear current-voltage (I-V) characteristics of FTJs into account, we propose equations for simulating the STDP curve parameters of a magnitude of the conductance change (ΔGmax) and a time window (τC) from the spike parameters of a peak amplitude (Vpeak) and time durations (tp and td) for three spike types: triangle-triangle, rectangular-triangle, and rectangular-rectangular. The power consumption experiments of the STDP revealed that the power consumption under the inactive-synapse condition (spike timing |Δt| > τC) was as large as 50–82% of that under the active-synapse condition (|Δt| < τC). This finding indicates that the power consumption under the inactive-synapse condition should be reduced to minimize the total power consumption of an SNN implemented by using FTJs as synapses.

Dendritic small conductance calcium-activated potassium channels activated by action potentials suppress EPSPs and gate spike-timing dependent synaptic plasticity.

Small conductance calcium-activated potassium channels (SK channels) are present in spines and can be activated by backpropagating action potentials (APs). This suggests they may play a critical role in spike-timing dependent synaptic plasticity (STDP). Consistent with this idea, EPSPs in both cortical and hippocampal pyramidal neurons were suppressed by preceding APs in an SK-dependent manner. In cortical pyramidal neurons EPSP suppression by preceding APs depended on their precise timing as well as the distance of activated synapses from the soma, was dendritic in origin, and involved SK-dependent suppression of NMDA receptor activation. As a result SK channel activation by backpropagating APs gated STDP induction during low-frequency AP-EPSP pairing, with both LTP and LTD absent under control conditions but present after SK channel block. These findings indicate that activation of SK channels in spines by backpropagating APs plays a key role in regulating both EPSP amplitude and STDP induction.

Adrenergic Gate Release for Spike Timing-Dependent Synaptic Potentiation

Spike timing-dependent synaptic plasticity (STDP) serves as a key cellular correlate of associative learning, which is facilitated by elevated attentional and emotional states involving activation of adrenergic signaling. At cellular levels, adrenergic signaling increases dendrite excitability, but the underlying mechanisms remain elusive. Here we show that activation of β2-adrenoceptors promoted STD long-term synaptic potentiation at mouse hippocampal excitatory synapses by inactivating dendritic Kv1.1-containingpotassium channels, which increased dendrite excitability and facilitated dendritic propagation of postsynaptic depolarization, potentially improving coincidental activation of pre- and postsynaptic terminals. We further demonstrate that adrenergic modulation of Kv1.1 was mediated by the signaling scaffold SAP97, which, through direct protein-protein interactions, escorts β2 signaling to remove Kv1.1 from the dendrite surface. These results reveal a mechanism through which the postsynaptic signaling scaffolds bridge the aroused brain state to promote induction of synaptic plasticity andpotentially to enhance spike timing and memory encoding.

Differential Regulation of Human Paired Associative Stimulation-Induced and Theta-Burst Stimulation-Induced Plasticity by L-type and T-type Ca2+ Channels.

Activity-dependent changes of postsynaptic Ca2+-concentration are influenced by a variety of different Ca2+-channels and play an important role in synaptic plasticity. Paired associative stimulation (PAS) and theta-burst stimulation (TBS) are noninvasive magnetic stimulation protocols used in human subjects to induce lasting corticospinal excitability changes that have been likened to synaptic long-term potentiation and long-term depression. To better characterize the Ca2+-related physiological mechanisms underlying PAS- and TBS-induced plasticity, we examined the impact of different Ca2+-sources. PAS-induced facilitation of corticospinal excitability was blocked by NMDA-receptor blocker dextromethorphan (DXM) and L-type voltage gated Ca2+ channels (VGCC) blocker nimodipine (NDP), but turned into depression by T-type VGCC blocker ethosuximide (ESM). Although, surprisingly, static corticospinal excitability was increased by the combination of DXM and NDP, PAS-induced facilitation was blocked. TBS-induced facilitation of corticospinal excitability, which has previously been shown to be turned into depression by L-type VGCC blocker NDP (Wankerl K, Weise D, Gentner R, Rumpf J, Classen J. 2010. L-type voltage-gated Ca2+ channels: a single molecular switch for long-term potentiation/long-term depression-like plasticity and activity-dependent metaplasticity in humans. J Neurosci. 30(18):6197-6204.), was blocked, but not reverted, by T-type VGCC blocker ESM. The different patterns of Ca2+-channel modulation of PAS- and TBS-induced plasticity may point to an important role of backpropagating action potentials in PAS-induced plasticity, similar as in spike-timing dependent synaptic plasticity, and to a requirement of dendritic Ca2+-dependent spikes in TBS-induced plasticity.

A Novel Hardware-Efficient Asynchronous Cellular Automaton Model of Spike-Timing-Dependent Synaptic Plasticity

The spike-timing-dependent synaptic plasticity (STDP) is a form of change of a synaptic weight depending on timings of pre- and postsynaptic spikes. In this brief, a novel asynchronous cellular automaton model of the STDP is presented. It is shown that the presented model can reproduce various STDP characteristics observed in physiological experiments. Also, the presented model is implemented in a field-programmable gate array, and laboratory experiments validate the reproductions of the physiological STDP characteristics. It is then shown that the presented model consumes much less hardware resources than a computationally efficient differential equation model of the STDP.

Oscillation, Conduction Delays, and Learning Cooperate to Establish Neural Competition in Recurrent Networks.

Specific memory might be stored in a subnetwork consisting of a small population of neurons. To select neurons involved in memory formation, neural competition might be essential. In this paper, we show that excitable neurons are competitive and organize into two assemblies in a recurrent network with spike timing-dependent synaptic plasticity (STDP) and axonal conduction delays. Neural competition is established by the cooperation of spontaneously induced neural oscillation, axonal conduction delays, and STDP. We also suggest that the competition mechanism in this paper is one of the basic functions required to organize memory-storing subnetworks into fine-scale cortical networks.

Network response synchronization enhanced by synaptic plasticity

Synchronization of neural network response on spatially localized periodic stimulation was studied. The network consisted of synaptically coupled spiking neurons with spike-timing-dependent synaptic plasticity (STDP). Network connectivity was defined by time evolving matrix of synaptic weights. We found that the steady-state spatial pattern of the weights could be rearranged due to locally applied external periodic stimulation. A method for visualization of synaptic weights as vector field was introduced to monitor the evolving connectivity matrix. We demonstrated that changes in the vector field and associated weight rearrangements underlay an enhancement of synchronization range.

A model of hippocampal spiking responses to items during learning of a context-dependent task.

Single unit recordings in the rat hippocampus have demonstrated shifts in the specificity of spiking activity during learning of a contextual item-reward association task. In this task, rats received reward for responding to different items dependent upon the context an item appeared in, but not dependent upon the location an item appears at. Initially, neurons in the rat hippocampus primarily show firing based on place, but as the rat learns the task this firing became more selective for items. We simulated this effect using a simple circuit model with discrete inputs driving spiking activity representing place and item followed sequentially by a discrete representation of the motor actions involving a response to an item (digging for food) or the movement to a different item (movement to a different pot for food). We implemented spiking replay in the network representing neural activity observed during sharp-wave ripple events, and modified synaptic connections based on a simple representation of spike-timing dependent synaptic plasticity. This simple network was able to consistently learn the context-dependent responses, and transitioned from dominant coding of place to a gradual increase in specificity to items consistent with analysis of the experimental data. In addition, the model showed an increase in specificity toward context. The increase of selectivity in the model is accompanied by an increase in binariness of the synaptic weights for cells that are part of the functional network.

Neuronal plasticity during sleep slow wave oscillations

Sleep is critical for regulation of synaptic efficacy, consolidation of memories and learning. It has been proposed that synaptic plasticity associated with sleep rhythms could contribute to consolidation of memories acquired during wakefulness. It was suggested that in the active state of sleep slow wave oscillation, the hippocampal formation activates latent memories stored in the neocortex (replay) and induce permanent changes in synaptic conductances. In this study, we present a thalamocortical network model of the slow-wave sleep activity characterized by repeatable (< 1 Hz) transitions between active (Up) and silent (Down) states of the network. The model consisted a layer of thalamic relay (TC) and reticular (RE) neurons in the thalamus as well as a model of cortical column with pyramidal neurons and inhibitory interneurons. All neurons were modeled based on the Hodgkin-Huxley kinetics. Spike-timing dependent synaptic plasticity was implemented to regulate synaptic efficacy. The slow oscillation was driven by intracortical dynamics; active states were initiated by the spontaneous miniature synaptic releases at different network sites [1] and propagated within the cortical network. The properties of slow oscillations depended on the intrinsic conductances, miniature synaptic releases, and the synaptic strength. In agreement with previous experimental studies, transition from the active to the silent state occurred more synchronously than the active state initiation. We found that the pattern of the active state propagation depended on the spatiotemporal pattern of activity on the previous cycle of slow oscillation and could be influenced by external input. Because of the refractoriness properties of the network, a probability of the next active state initiation was higher for the network site that initiated activity at the previous cycle. Furthermore, even weak external stimulation delivered to the network results in increase probability of up-state induction at the stimulation location. This suggests that spatially and temporally sparse hippocampal input could influence the spatiotemporal pattern of slow oscillation. Location of the initiation site and the pattern of active state propagation determined the relative timing of spiking in cortical neurons. When spike-time dependent synaptic plasticity (STDP) was implemented, there was a net decrease in synaptic strengths and, at the same time, an increase in the strength of specific synapses which were associated with the sequence replay. The change in synaptic weights between any two neurons was determined by the direction of active state propagation and by the distance between the neurons. Our study propose a mechanism of how interaction between cortically generated slow waves and sparse external input, possibly representing input from hippocampal formation, may lead to reorganization of synaptic strength during stage 3/4 sleep.

Mechanisms Underlying the Rules for Associative Plasticity at Adult Human Neocortical Synapses

The neocortex in our brain stores long-term memories by changing the strength of connections between neurons. To date, the rules and mechanisms that govern activity-induced synaptic changes at human cortical synapses are poorly understood and have not been studied directly at a cellular level. Here, we made whole-cell recordings of human pyramidal neurons in slices of brain tissue resected during neurosurgery to investigate spike timing-dependent synaptic plasticity in the adult human neocortex. We find that human cortical synapses can undergo bidirectional modifications in strength throughout adulthood. Both long-term potentiation and long-term depression of synapses was dependent on postsynaptic NMDA receptors. Interestingly, we find that human cortical synapses can associate presynaptic and postsynaptic events in a wide temporal window, and that rules for synaptic plasticity in human neocortex are reversed compared with what is generally found in the rodent brain. We show this is caused by dendritic L-type voltage-gated Ca2+ channels that are prominently activated during action potential firing. Activation of these channels determines whether human synapses strengthen or weaken. These findings provide a synaptic basis for the timing rules observed in human sensory and motor plasticity in vivo, and offer insights into the physiological role of L-type voltage-gated Ca2+ channels in the human brain.

Microcircuit mechanisms involved in paired associative stimulation‐induced depression of corticospinal excitability

Synaptic weight changes induced by temporal correlations between the spikes of pre- and postsynaptic neurons are referred to as spike-timing-dependent plasticity (STDP). Transcranial magnetic stimulation (TMS) induces long-lasting effects on corticospinal excitability, if it is repetitively paired with stimulation of afferents from a corresponding contralateral hand region at short intervals (paired associative stimulation, PAS). PAS-induced plasticity has been linked with synaptic STDP. We aimed to investigate which elements of the cortical microcircuitry sustain and govern PAS-induced depression of corticospinal excitability in the target muscle representation (and enhancement of excitability in its functional surround). We show that the time window during which the interaction between both stimulus-induced cortical events leads to immediate post-interventional depression is short (<4.5 ms). The depressant PAS effects at the target representation were completely blocked by applying a subthreshold magnetic pulse 3 ms before the principal TMS pulse, even when the strength of the latter was adjusted to generate a motor-evoked potential of similar amplitude to that with the unconditioned magnetic pulse. Epidural recordings from the cervical cord of a patient showed that under this condition late TMS-evoked I-waves remain suppressed. When the intensity of the TMS component during PAS was lowered - sufficient to allow activation of inhibitory neurons, but insufficient to activate corticospinal neurons - excitability of short-latency intracortical inhibition remained unchanged. PAS-induced facilitation in the functional surround followed the same pattern as the centre-depressant effects. These findings may suggest that excitability-depressant PAS-induced effects are due to weakening of excitatory synapses between upper cortical layer principal neurons, but not those located on the corticospinal neuron, or inhibitory synapses. Inhibitory interneurons involved in short-latency intracortical inhibition are gate-keepers to producing centre-depressant/surround-facilitatory PAS effects. Based on these and earlier findings we propose a model specifying the composition and laminar location of the involved microcircuit of PAS-induced plasticity that may enhance its utility as a model of STDP in humans.

Modulation of synaptic plasticity by the coactivation of spatially distinct synaptic inputs in rat hippocampal CA1 apical dendrites

The phenomenon whereby the relative timing between presynaptic and postsynaptic spiking determines the direction and extent of synaptic changes in a critical temporal window is known as spike timing-dependent synaptic plasticity (STDP). We have previously reported that STDP profiles can be classified into two types depending on their layer-specific location along CA1 pyramidal neuron dendrites in the rat hippocampus, suggesting that there are differences in information processing between the proximal dendrite (PD) and distal dendrite (DD). However, how the different types of information processing interact at different dendritic locations remains unclear. To investigate how the temporal information of inputs to PD influences information processing at DD, PD stimulation was applied while the STDP protocol was simultaneously applied at DDs of CA1 pyramidal neurons. Synaptic plasticity induced by the STDP protocol at DDs was enhanced or depressed depending on the timing of the back-propagating action potentials (bAPs) and the excitatory and inhibitory postsynaptic potentials elicited by PD stimulation. These results suggested that bAPs function as carriers of temporal information of PD inputs to DD. Next, the influence of DD on PD was investigated using the same protocol. Synaptic plasticity at PD was modulated only if the pairing stimuli were applied to elicit coincidental timing of bAP and the excitatory postsynaptic potential. Such coding modulations could provide the basis for a novel learning rule and may be important factors in the integration of spatiotemporal input information in neural networks in the brain.

Dynamic Observation of Brain-Like Learning in a Ferroelectric Synapse Device

A brain-like learning function was implemented in an electronic synapse device using a ferroelectric-gate field effect transistor (FeFET). The FeFET was a bottom-gate type FET with a ZnO channel and a ferroelectric Pb(Zr,Ti)O3 (PZT) gate insulator. The synaptic weight, which is represented by the channel conductance of the FeFET, is updated by applying a gate voltage through a change in the ferroelectric polarization in the PZT. A learning function based on the symmetric spike-timing dependent synaptic plasticity was implemented in the synapse device using the multilevel weight update by applying a pulse gate voltage. The dynamic weighting and learning behavior in the synapse device was observed as a change in the membrane potential in a spiking neuron circuit.