non-Hebbian Plasticity Research Articles

Dendritic, delayed, stochastic CaMKII activation in behavioural time scale plasticity

Behavioural time scale plasticity (BTSP) is non-Hebbian plasticity induced by integrating presynaptic and postsynaptic components separated by a behaviourally relevant time scale (seconds)1. BTSP in hippocampal CA1 neurons underlies place cell formation. However, the molecular mechanisms that enable synapse-specific plasticity on a behavioural time scale are unknown. Here we show that BTSP can be induced in a single dendritic spine using two-photon glutamate uncaging paired with postsynaptic current injection temporally separated by a behavioural time scale. Using an improved Ca2+/calmodulin-dependent kinase II (CaMKII) sensor, we did not detect CaMKII activation during this BTSP induction. Instead, we observed dendritic, delayed and stochastic CaMKII activation (DDSC) associated with Ca2+ influx and plateau potentials 10–100 s after BTSP induction. DDSC required both presynaptic and postsynaptic activity, which suggests that CaMKII can integrate these two signals. Also, optogenetically blocking CaMKII 15–30 s after the BTSP protocol inhibited synaptic potentiation, which indicated that DDSC is an essential mechanism of BTSP. IP3-dependent intracellular Ca2+ release facilitated both DDSC and BTSP. Thus, our study suggests that non-synapse-specific CaMKII activation provides an instructive signal with an extensive time window over tens of seconds during BTSP.

Non-Hebbian plasticity transforms transient experiences into lasting memories.

The dominant models of learning and memory, such as Hebbian plasticity, propose that experiences are transformed into memories through input-specific synaptic plasticity at the time of learning. However, synaptic plasticity is neither strictly input-specific nor restricted to the time of its induction. The impact of such forms of non-Hebbian plasticity on memory has been difficult to test, and hence poorly understood. Here, we demonstrate that synaptic manipulations can deviate from the Hebbian model of learning, yet produce a lasting memory. First, we established a weak associative conditioning protocol in mice, where optogenetic stimulation of sensory thalamic input to the amygdala was paired with a footshock, but no detectable memory was formed. However, when the same input was potentiated minutes before or after, or even 24 hr later, the associative experience was converted into a lasting memory. Importantly, potentiating an independent input to the amygdala minutes but not 24 hr after the pairing produced a lasting memory. Thus, our findings suggest that the process of transformation of a transient experience into a memory is neither restricted to the time of the experience nor to the synapses triggered by it; instead, it can be influenced by past and future events.

Non-Hebbian plasticity transforms transient experiences into lasting memories

Non-Hebbian plasticity transforms transient experiences into lasting memories

Astrocyte D1/D5 Dopamine Receptors Govern Non-Hebbian Long-Term Potentiation at Sensory Synapses onto Lamina I Spinoparabrachial Neurons.

Recent work demonstrated that activation of spinal D1 and D5 dopamine receptors (D1/D5Rs) facilitates non-Hebbian long-term potentiation (LTP) at primary afferent synapses onto spinal projection neurons. However, the cellular localization of the D1/D5Rs driving non-Hebbian LTP in spinal nociceptive circuits remains unknown, and it is also unclear whether D1/D5R signaling must occur concurrently with sensory input in order to promote non-Hebbian LTP at these synapses. Here we investigate these issues using cell-type-selective knockdown of D1Rs or D5Rs from lamina I spinoparabrachial neurons, dorsal root ganglion (DRG) neurons, or astrocytes in adult mice of either sex using Cre recombinase-based genetic strategies. The LTP evoked by low-frequency stimulation of primary afferents in the presence of the selective D1/D5R agonist SKF82958 persisted following the knockdown of D1R or D5R in spinoparabrachial neurons, suggesting that postsynaptic D1/D5R signaling was dispensable for non-Hebbian plasticity at sensory synapses onto these key output neurons of the superficial dorsal horn (SDH). Similarly, the knockdown of D1Rs or D5Rs in DRG neurons failed to influence SKF82958-enabled LTP in lamina I projection neurons. In contrast, SKF82958-induced LTP was suppressed by the knockdown of D1R or D5R in spinal astrocytes. Furthermore, the data indicate that the activation of D1R/D5Rs in spinal astrocytes can either retroactively or proactively drive non-Hebbian LTP in spinoparabrachial neurons. Collectively, these results suggest that dopaminergic signaling in astrocytes can strongly promote activity-dependent LTP in the SDH, which is predicted to significantly enhance the amplification of ascending nociceptive transmission from the spinal cord to the brain.

Normative Framework for Deriving Neural Networks with Multicompartmental Neurons and Non-Hebbian Plasticity

Normative Framework for Deriving Neural Networks with Multicompartmental Neurons and Non-Hebbian Plasticity

D1/D5 Dopamine Receptors and mGluR5 Jointly Enable Non-Hebbian Long-Term Potentiation at Sensory Synapses onto Lamina I Spinoparabrachial Neurons

Highly correlated firing of primary afferent inputs and lamina I projection neurons evokes synaptic long-term potentiation (LTP), a mechanism by which ascending nociceptive transmission can be amplified at the level of the spinal dorsal horn. However, the degree to which neuromodulatory signaling shapes the temporal window governing spike-timing-dependent plasticity (STDP) at sensory synapses onto projection neurons remains unclear. The present study demonstrates that activation of spinal D1/D5 dopamine receptors (D1/D5Rs) creates a highly permissive environment for the production of LTP in male and female adult mouse spinoparabrachial neurons by promoting non-Hebbian plasticity. Bath application of the mixed D1/D5R agonist SKF82958 unmasked LTP at STDP pairing intervals that normally fail to alter synaptic efficacy. Furthermore, during D1/D5R signaling, action potential discharge in projection neurons became dispensable for LTP generation, and primary afferent stimulation alone was sufficient to induce strengthening of sensory synapses. This non-Hebbian LTP was blocked by the D1/D5R antagonist SCH 39166 or genetic deletion of D5R, and required activation of mGluR5 and intracellular Ca2+ release but was independent of NMDAR activation. D1/D5R-enabled non-Hebbian plasticity was observed across multiple neuronal subpopulations in the superficial dorsal horn but was more prevalent in spinoparabrachial neurons than interneurons. Interestingly, the ability of neonatal tissue damage to promote non-Hebbian LTP in adult projection neurons was not observed in D5R knock-out mice. Collectively, these findings suggest that joint spinal D1/D5R and mGluR5 activation can allow unfettered potentiation of sensory synapses onto the output neurons responsible for conveying pain and itch information to the brain.SIGNIFICANCE STATEMENT Synaptic LTP in spinal projection neurons has been implicated in the generation of chronic pain. Under normal conditions, plasticity at sensory synapses onto adult mouse spinoparabrachial neurons follows strict Hebbian learning rules, requiring coincident presynaptic and postsynaptic firing. Here, we demonstrate that the activation of spinal D1/D5Rs promotes a switch from Hebbian to non-Hebbian LTP so that primary afferent stimulation alone is sufficient to evoke LTP in the absence of action potential discharge in projection neurons, which required joint activation of mGluR5 and intracellular Ca2+ release but not NMDARs. These results suggest that D1/D5Rs cooperate with mGluR5 receptors in the spinal dorsal horn to powerfully influence the amplification of ascending nociceptive transmission to the brain.

A Biologically Plausible Neural Network for Multichannel Canonical Correlation Analysis.

Cortical pyramidal neurons receive inputs from multiple distinct neural populations and integrate these inputs in separate dendritic compartments. We explore the possibility that cortical microcircuits implement canonical correlation analysis (CCA), an unsupervised learning method that projects the inputs onto a common subspace so as to maximize the correlations between the projections. To this end, we seek a multichannel CCA algorithm that can be implemented in a biologically plausible neural network. For biological plausibility, we require that the network operates in the online setting and its synaptic update rules are local. Starting from a novel CCA objective function, we derive an online optimization algorithm whose optimization steps can be implemented in a single-layer neural network with multicompartmental neurons and local non-Hebbian learning rules. We also derive an extension of our online CCA algorithm with adaptive output rank and output whitening. Interestingly, the extension maps onto a neural network whose neural architecture and synaptic updates resemble neural circuitry and non-Hebbian plasticity observed in the cortex.

The Complex Formed by Group I Metabotropic Glutamate Receptor (mGluR) and Homer1a Plays a Central Role in Metaplasticity and Homeostatic Synaptic Scaling.

G-protein-coupled receptors can be constitutively activated following physical interaction with intracellular proteins. The first example described was the constitutive activation of Group I metabotropic glutamate receptors (mGluR: mGluR1,5) following their interaction with Homer1a, an activity-inducible early-termination variant of the scaffolding protein Homer that lacks dimerization capacity (Ango et al., 2001). Homer1a disrupts the links, maintained by the long form of Homer (cross-linking Homers), between mGluR1,5 and the Shank-GKAP-PSD-95-ionotropic glutamate receptor network. Two characteristics of the constitutive activation of the Group I mGluR-Homer1a complex are particularly interesting: (1) it affects a large number of synapses in which Homer1a is upregulated following enhanced, long-lasting neuronal activity; and (2) it mainly depends on Homer1a protein turnover. The constitutively active Group I mGluR-Homer1a complex is involved in the two main forms of non-Hebbian neuronal plasticity: "metaplasticity" and "homeostatic synaptic scaling," which are implicated in a large series of physiological and pathologic processes. Those include non-Hebbian plasticity observed in visual system, synapses modulated by addictive drugs (rewarded synapses), chronically overactivated synaptic networks, normal sleep, and sleep deprivation.

Learning to Generate Sequences with Combination of Hebbian and Non-hebbian Plasticity in Recurrent Spiking Neural Networks.

Synaptic Plasticity, the foundation for learning and memory formation in the human brain, manifests in various forms. Here, we combine the standard spike timing correlation based Hebbian plasticity with a non-Hebbian synaptic decay mechanism for training a recurrent spiking neural model to generate sequences. We show that inclusion of the adaptive decay of synaptic weights with standard STDP helps learn stable contextual dependencies between temporal sequences, while reducing the strong attractor states that emerge in recurrent models due to feedback loops. Furthermore, we show that the combined learning scheme suppresses the chaotic activity in the recurrent model substantially, thereby enhancing its' ability to generate sequences consistently even in the presence of perturbations.

A different form of synaptic plasticity

Neuroscience How do synaptic or other neuronal changes support learning? This subject has been dominated by Hebb's postulate of synaptic change. Although there is strong experimental support for Hebbian plasticity in a number of preparations, alternative ideas have also been developed over the years. Bittner et al. provide in vivo, in vitro, and modeling data to support the view that non-Hebbian plasticity may underlie the formation of hippocampal place fields (see the Perspective by Krupic). Instead of multiple pairings, a single strong Ca2+ plateau potential in neuronal dendrites paired with spatial inputs may be sufficient to produce place cells. Science , this issue p. [1033][1]; see also p. [974][2] [1]: /lookup/doi/10.1126/science.aan3846 [2]: /lookup/doi/10.1126/science.aao4159

Burst-Dependent Bidirectional Plasticity in the Cerebellum Is Driven by Presynaptic NMDA Receptors.

Numerous studies have shown that cerebellar function is related to the plasticity at the synapses between parallel fibers and Purkinje cells. How specific input patterns determine plasticity outcomes, as well as the biophysics underlying plasticity of these synapses, remain unclear. Here, we characterize the patterns of activity that lead to postsynaptically expressed LTP using both in vivo and in vitro experiments. Similar to the requirements of LTD, we find that high-frequency bursts are necessary to trigger LTP and that this burst-dependent plasticity depends on presynaptic NMDA receptors and nitric oxide (NO) signaling. We provide direct evidence for calcium entry through presynaptic NMDA receptors in a subpopulation of parallel fiber varicosities. Finally, we develop and experimentally verify a mechanistic plasticity model based on NO and calcium signaling. The model reproduces plasticity outcomes from data and predicts the effect of arbitrary patterns of synaptic inputs on Purkinje cells, thereby providing a unified description of plasticity.

Long-term Synaptic Plasticity: Circuit Perturbation and Stabilization.

At central synapses, activity-dependent synaptic plasticity has a crucial role in information processing, storage, learning, and memory under both physiological and pathological conditions. One widely accepted model of learning mechanism and information processing in the brain is Hebbian Plasticity: long-term potentiation (LTP) and long-term depression (LTD). LTP and LTD are respectively activity-dependent enhancement and reduction in the efficacy of the synapses, which are rapid and synapse-specific processes. A number of recent studies have a strong focal point on the critical importance of another distinct form of synaptic plasticity, non-Hebbian plasticity. Non-Hebbian plasticity dynamically adjusts synaptic strength to maintain stability. This process may be very slow and occur cell-widely. By putting them all together, this mini review defines an important conceptual difference between Hebbian and non-Hebbian plasticity.

Non-Hebbian plasticity at C-fiber synapses in rat spinal cord lamina I neurons

Current concepts of memory storage are largely based on Hebbian-type synaptic long-term potentiation induced by concurrent activity of pre- and postsynaptic neurons. Little is known about non-Hebbian synaptic plasticity, which, if present in nociceptive pathways, could resolve a number of unexplained findings. We performed whole-cell patch-clamp recordings in rat spinal cord slices and found that a rise in postsynaptic [Ca2+]i due to postsynaptic depolarization was sufficient to induce synaptic long-term potentiation (LTP) in the absence of any presynaptic conditioning stimulation. LTP induction could be prevented by postsynaptic application of the Ca2+ chelator BAPTA (1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid), the L-type voltage-gated calcium channel (VGCC) antagonist nifedipine, and by postsynaptic application of the NMDA receptor antagonist MK801. This indicates that synaptic potentiation was induced postsynaptically by Ca2+ entry likely via L-type voltage-gated Ca2+ channels (VGCC) and via NMDA receptor channels. The paired pulse ratio and the coefficient of variation remained unchanged in neurons expressing LTP, suggesting that this form of synaptic potentiation was not only induced, but also expressed postsynaptically. Postsynaptic depolarization had no influence on firing patterns, action potential shape, or neuronal excitability. An increase in [Ca2+]i in spinal lamina I neurons induces a non-Hebbian form of synaptic plasticity in spinal nociceptive pathways without affecting neuronal active and passive membrane properties.

Homeostatic Scaling Requires Group I mGluR Activation Mediated by Homer1a

Homeostatic scaling is a non-Hebbian form of neural plasticity that maintains neuronal excitability and informational content of synaptic arrays in the face of changes of network activity. Here, we demonstrate that homeostatic scaling is dependent on group I metabotropic glutamate receptor activation that is mediated by the immediate early gene Homer1a. Homer1a is transiently upregulated during increases in network activity and evokes agonist-independent signaling of group I mGluRs that scales down the expression of synaptic AMPA receptors. Homer1a effects are dynamic and play a role in the induction of scaling. Similar to mGluR-LTD, Homer1a-dependent scaling involves a reduction of tyrosine phosphorylation of GluA2 (GluR2), but is distinct in that it exploits a unique signaling property of group I mGluR to confer cell-wide, agonist-independent activation of the receptor. These studies reveal an elegant interplay of mechanisms that underlie Hebbian and non-Hebbian plasticity.

Non-Hebbian Synaptic Plasticity Induced by Repetitive Postsynaptic Action Potentials

Modern theories on memory storage have mainly focused on Hebbian long-term potentiation (LTP), which requires coincident activation of presynaptic and postsynaptic neurons for its induction. In addition to Hebbian LTP, the roles of non-Hebbian plasticity have also been predicted by some neuronal network models. However, still only a few pieces of evidence have been presented for the presence of such plasticity. In this study, we show in mouse hippocampal slices that LTP can be induced by postsynaptic repetitive depolarization alone in the absence of presynaptic inputs. The induction was dependent on voltage-dependent calcium channels instead of NMDA receptors (NMDARs), whereas the expression mechanism was shared with conventional NMDAR-dependent LTP. During the potentiation, the amplitude of spontaneous EPSCs was increased, suggesting a novel neuron-wide nature of this form of LTP. Furthermore, we also successfully induced LTP with trains of action potentials, which supported the possible existence of depolarizing pulse-induced LTP in vivo. Based on these findings, we suggest a model in which neuron-wide LTP works in concert with synapse-specific Hebbian plasticity to help information processing in memory formation.

A Genetic Switch for Epilepsy in Adult Mice

Premature death from seizures afflicts gene-targeted mice expressing the Q/R site-unedited glutamate receptor subunit GluR-B(Q) of AMPA receptors in central neurons. Early seizure-related death has now been circumvented by a genetic switch that restricts GluR-B(Q) expression to forebrain principal neurons from postnatal stages onward, prominently in hippocampus and striatum and less so in cortex and amygdala. When switched on, functional receptor incorporation of GluR-B(Q) could be demonstrated by imaging evoked AMPA channel-mediated spinous Ca2+ transients in CA1 pyramidal cells. Sustained GluR-B(Q) expression in adult mice led to smaller excitatory postsynaptic responses in the CA1 region with unchanged presynaptic fiber excitability. Notably, despite the smaller excitatory response, the CA1 cells exhibited a reduced population spike threshold, which might underlie the spontaneous manifestations of epilepsy, including myocloni and generalized seizures with limbic components, observed by synchronous video monitoring and electroencephalographic recordings. No neuropathological symptoms developed when GluR-B(Q) expression was restricted to only hippocampal neurons. Our results show that seizure susceptibility is triggered by GluR-B(Q) expression also in the adult brain and that circuit hyperexcitability is not an immediate consequence of GluR-B(Q) but requires yet unknown downstream events, likely to be induced by non-Hebbian plasticity from Ca2+-permeable AMPA channels in principal neurons.