Partial Breakdown of Input Specificity of STDP at Individual Synapses Promotes New Learning

Half a century legacy of long-term potentiation

In 1973, two papers from Bliss and Lømo and from Bliss and Gardner-Medwin reported that high-frequency synaptic stimulation in the dentate gyrus of rabbits resulted in a long-lasting increase in synaptic strength. This form of synaptic plasticity, commonly referred to as long-term potentiation (LTP), was immediately considered as an attractive mechanism accounting for the ability of the brain to store information. In this historical piece looking back over the past 50 years, we discuss how these two landmark contributions directly motivated a colossal research effort and detail some of the resulting milestones that have shaped our evolving understanding of the molecular and cellular underpinnings of LTP. We highlight the main features of LTP, cover key experiments that defined its induction and expression mechanisms, and outline the evidence supporting a potential role of LTP in learning and memory. We also briefly explore some ramifications of LTP on network stability, consider current limitations of LTP as a model of associative memory, and entertain future research orientations.

Cdc42 activation is necessary for heterosynaptic cooperation and competition

Synapses change their weights in response to neuronal activity and in turn, neuronal networks alter their response properties and ultimately allow the brain to store information as memories. As for memories, not all events are maintained over time. Maintenance of synaptic plasticity depends on the interplay between functional changes at synapses and the synthesis of plasticity-related proteins that are involved in stabilizing the initial functional changes. Different forms of synaptic plasticity coexist in time and across the neuronal dendritic area. Thus, homosynaptic plasticity refers to activity-dependent synaptic modifications that are input-specific, whereas heterosynaptic plasticity relates to changes in non-activated synapses. Heterosynaptic forms of plasticity, such as synaptic cooperation and competition allow neurons to integrate events that occur separated by relatively large time windows, up to one hour. Here, we show that activation of Cdc42, a Rho GTPase that regulates actin cytoskeleton dynamics, is necessary for the maintenance of long-term potentiation (LTP) in a time-dependent manner. Inhibiting Cdc42 activation does not alter the time-course of LTP induction and its initial expression but blocks its late maintenance. We show that Cdc42 activation is involved in the phosphorylation of cofilin, a protein involved in modulating actin filaments and that weak and strong synaptic activation leads to similar levels on cofilin phosphorylation, despite different levels of LTP expression. We show that Cdc42 activation is required for synapses to interact by cooperation or competition, supporting the hypothesis that modulation of the actin cytoskeleton provides an activity-dependent and time-restricted permissive state of synapses allowing synaptic plasticity to occur. We found that under competition, the sequence in which synapses are activated determines the degree of LTP destabilization, demonstrating that competition is an active destabilization process. Taken together, we show that modulation of actin cytoskeleton by Cdc42 activation is necessary for the expression of homosynaptic and heterosynaptic forms of plasticity. Determining the temporal and spatial rules that determine whether synapses cooperate or compete will allow us to understand how memories are associated.

Plasticity of Response Properties of Mouse Visual Cortex Neurons Induced by Optogenetic Tetanization In Vivo

Heterosynaptic plasticity, along with Hebbian homosynaptic plasticity, is an important mechanism ensuring the stable operation of learning neuronal networks. However, whether heterosynaptic plasticity occurs in the whole brain in vivo, and what role(s) in brain function in vivo it could play, remains unclear. Here, we used an optogenetics approach to apply a model of intracellular tetanization, which was established and employed to study heterosynaptic plasticity in brain slices, to study the plasticity of response properties of neurons in the mouse visual cortex in vivo. We show that optogenetically evoked high-frequency bursts of action potentials (optogenetic tetanization) in the principal neurons of the visual cortex induce long-term changes in the responses to visual stimuli. Optogenetic tetanization had distinct effects on responses to different stimuli, as follows: responses to optimal and orthogonal orientations decreased, responses to null direction did not change, and responses to oblique orientations increased. As a result, direction selectivity of the neurons decreased and orientation tuning became broader. Since optogenetic tetanization was a postsynaptic protocol, applied in the absence of sensory stimulation, and, thus, without association of presynaptic activity with bursts of action potentials, the observed changes were mediated by mechanisms of heterosynaptic plasticity. We conclude that heterosynaptic plasticity can be induced in vivo and propose that it may play important homeostatic roles in operation of neural networks by helping to prevent runaway dynamics of responses to visual stimuli and to keep the tuning of neuronal responses within the range optimized for the encoding of multiple features in population activity.

Ethanol Differentially Affects Excitatory and Inhibitory Synaptic Transmission in Visual Cortex of Wild-type and Adenosine A1R Knock-out Mice

Ethanol is one of the most commonly used and abused substances in the world. While the behavioral effects of ethanol are well characterized, mechanisms of its action on neurons and synapses remain elusive. Prior research suggested that ethanol could affect neurons by interfering with metabolism of biologically active molecules, such as adenosine. Here, we explored the involvement of adenosine A1 receptors (A1R) in mediating ethanol's effects on synaptic transmission to layer 2/3 pyramidal neurons of visual cortex using wild type (WT) and A1R knock-out (KO) mice. Ethanol differentially affected excitatory and inhibitory transmission in WT and KO mice. In slices from WT mice ethanol had heterogeneous effects on excitatory transmission (facilitation, suppression or no change), with no net change. Ethanol's effects remained heterogeneous during acute blockade of A1Rs with a selective antagonist DPCPX. However, in A1RKO mice ethanol consistently suppressed excitatory transmission, with no cases of enhancement observed. Inhibitory transmission was suppressed by ethanol in both WT and A1RKO mice. At both excitatory and inhibitory synapses, changes of response amplitude correlated with changes of paired-pulse ratio, suggesting involvement of presynaptic mechanisms. We conclude that A1Rs are not involved in mediating effects of ethanol on synaptic transmission in mouse visual cortex. However, A1Rs are necessary for development of mechanisms mediating facilitation at some excitatory synapses. Our results add evidence for the diversity of ethanol's effects and mechanisms of action on synaptic transmission in different brain structures, and even in the same brain area (visual cortex) in different species, rats vs mice.

Interaction between Theta Phase and Spike Timing-Dependent Plasticity Simulates Theta-Induced Memory Effects

Rodent studies suggest that spike timing relative to hippocampal theta activity determines whether potentiation or depression of synapses arise. Such changes also depend on spike timing between presynaptic and postsynaptic neurons, known as spike timing-dependent plasticity (STDP). STDP, together with theta phase-dependent learning, has inspired several computational models of learning and memory. However, evidence to elucidate how these mechanisms directly link to human episodic memory is lacking. In a computational model, we modulate long-term potentiation (LTP) and long-term depression (LTD) of STDP, by opposing phases of a simulated theta rhythm. We fit parameters to a hippocampal cell culture study in which LTP and LTD were observed to occur in opposing phases of a theta rhythm. Further, we modulated two inputs by cosine waves with 0° and asynchronous phase offsets and replicate key findings in human episodic memory. Learning advantage was found for the in-phase condition, compared with the out-of-phase conditions, and was specific to theta-modulated inputs. Importantly, simulations with and without each mechanism suggest that both STDP and theta phase-dependent plasticity are necessary to replicate the findings. Together, the results indicate a role for circuit-level mechanisms, which bridge the gap between slice preparation studies and human memory.

Stability and learning in excitatory synapses by nonlinear inhibitory plasticity

Synaptic changes are hypothesized to underlie learning and memory formation in the brain. But Hebbian synaptic plasticity of excitatory synapses on its own is unstable, leading to either unlimited growth of synaptic strengths or silencing of neuronal activity without additional homeostatic mechanisms. To control excitatory synaptic strengths, we propose a novel form of synaptic plasticity at inhibitory synapses. Using computational modeling, we suggest two key features of inhibitory plasticity, dominance of inhibition over excitation and a nonlinear dependence on the firing rate of postsynaptic excitatory neurons whereby inhibitory synaptic strengths change with the same sign (potentiate or depress) as excitatory synaptic strengths. We demonstrate that the stable synaptic strengths realized by this novel inhibitory plasticity model affects excitatory/inhibitory weight ratios in agreement with experimental results. Applying a disinhibitory signal can gate plasticity and lead to the generation of receptive fields and strong bidirectional connectivity in a recurrent network. Hence, a novel form of nonlinear inhibitory plasticity can simultaneously stabilize excitatory synaptic strengths and enable learning upon disinhibition.

Enhanced Non-Associative Long-Term Potentiation in Immature Granule Cells in the Dentate Gyrus of Adult Rats

The dentate gyrus is one of the few sites of neurogenesis in the adult brain. Integration of new-generated granule cells into the hippocampal circuitry provides a substrate for structural plasticity, fundamental for normal function of adult hippocampus. However, mechanisms of synaptic plasticity that mediate integration of new-generated granule cells into the existing circuitry remain poorly understood. Especially mechanisms of plasticity at GABA-ergic synapses remain elusive. Here, we show that postsynaptic spiking without presynaptic activation can induce heterosynaptic, non-associative plasticity at GABA-ergic inputs to both immature and mature granule cells. In both immature and mature neurons, plastic changes were bidirectional and individual inputs could express long-term potentiation (LTP) or long-term depression (LTD), or do not change. However, properties of non-associative plasticity dramatically change with maturation of newly generated granule cells: while in immature cells there was a clear predominance of non-associative LTP and net potentiation across the inputs, in mature neurons, potentiation and depression were balanced with no net change on average. We conclude that GABA-ergic inputs to granule cells are plastic, and that the rules for induction of non-associative plasticity change with maturation. We propose that potentiation-biased non-associative plasticity of GABA-ergic transmission might help to counter-balance an increase of excitatory drive that is facilitated by enhanced LTP at glutamatergic synapses in maturating granule cells. Such mechanism might help to build a strong GABA-ergic input to surviving active new cells, necessary for normal function of mature granule cells, which operate under a tight inhibitory control and generate sparse spiking activity.

Detection of tACS Entrainment Critically Depends on Epoch Length

Neural entrainment is the phase synchronization of a population of neurons to an external rhythmic stimulus such as applied in the context of transcranial alternating current stimulation (tACS). tACS can cause profound effects on human behavior. However, there remain a significant number of studies that find no behavioral effect when tACS is applied to human subjects. To investigate this discrepancy, we applied time sensitive phase lock value (PLV) based analysis to single unit data from the rat motor cortex. The analysis revealed that detection of neural entrainment depends critically on the epoch length within which spiking information is accumulated. Increasing the epoch length allowed for detection of progressively weaker levels of neural entrainment. Based on this single unit analysis, we hypothesized that tACS effects on human behavior would be more easily detected in a behavior paradigm which utilizes longer epoch lengths. We tested this by using tACS to entrain tremor in patients and healthy volunteers. When the behavioral data were analyzed using short duration epochs tremor entrainment effects were not detectable. However, as the epoch length was progressively increased, weak tremor entrainment became detectable. These results suggest that tACS behavioral paradigms that rely on the accumulation of information over long epoch lengths will tend to be successful at detecting behavior effects. However, tACS paradigms that rely on short epoch lengths are less likely to detect effects.

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).

Altered Heterosynaptic Plasticity Impairs Visual Discrimination Learning in Adenosine A1 Receptor Knock-Out Mice

Theoretical and modeling studies demonstrate that heterosynaptic plasticity-changes at synapses inactive during induction-facilitates fine-grained discriminative learning in Hebbian-type systems, and helps to achieve a robust ability for repetitive learning. A dearth of tools for selective manipulation has hindered experimental analysis of the proposed role of heterosynaptic plasticity in behavior. Here we circumvent this obstacle by testing specific predictions about the behavioral consequences of the impairment of heterosynaptic plasticity by experimental manipulations to adenosine A1 receptors (A1Rs). Our prior work demonstrated that the blockade of adenosine A1 receptors impairs heterosynaptic plasticity in brain slices and, when implemented in computer models, selectively impairs repetitive learning on sequential tasks. Based on this work, we predict that A1R knock-out (KO) mice will express (1) impairment of heterosynaptic plasticity and (2) behavioral deficits in learning on sequential tasks. Using electrophysiological experiments in slices and behavioral testing of animals of both sexes, we show that, compared with wild-type controls, A1R KO mice have impaired synaptic plasticity in visual cortex neurons, coupled with significant deficits in visual discrimination learning. Deficits in A1R knockouts were seen specifically during relearning, becoming progressively more apparent with learning on sequential visual discrimination tasks of increasing complexity. These behavioral results confirm our model predictions and provide the first experimental evidence for a proposed role of heterosynaptic plasticity in organism-level learning. Moreover, these results identify heterosynaptic plasticity as a new potential target for interventions that may help to enhance new learning on a background of existing memories.SIGNIFICANCE STATEMENT Understanding how interacting forms of synaptic plasticity mediate learning is fundamental for neuroscience. Theory and modeling revealed that, in addition to Hebbian-type associative plasticity, heterosynaptic changes at synapses that were not active during induction are necessary for stable system operation and fine-grained discrimination learning. However, lacking tools for selective manipulation prevented behavioral analysis of heterosynaptic plasticity. Here we circumvent this barrier: from our prior experimental and computational work we predict differential behavioral consequences of the impairment of Hebbian-type versus heterosynaptic plasticity. We show that, in adenosine A1 receptor knock-out mice, impaired synaptic plasticity in visual cortex neurons is coupled with specific deficits in learning sequential, increasingly complex visual discrimination tasks. This provides the first evidence linking heterosynaptic plasticity to organism-level learning.

The Sync-Fire/deSync model: Modelling the reactivation of dynamic memories from cortical alpha oscillations

We propose a neural network model to explore how humans can learn and accurately retrieve temporal sequences, such as melodies, movies, or other dynamic content. We identify target memories by their neural oscillatory signatures, as shown in recent human episodic memory paradigms. Our model comprises three plausible components for the binding of temporal content, where each component imposes unique limitations on the encoding and representation of that content. A cortical component actively represents sequences through the disruption of an intrinsically generated alpha rhythm, where a desynchronisation marks information-rich operations as the literature predicts. A binding component converts each event into a discrete index, enabling repetitions through a sparse encoding of events. A timing component - consisting of an oscillatory "ticking clock" made up of hierarchical synfire chains - discretely indexes a moment in time. By encoding the absolute timing between discretised events, we show how one can use cortical desynchronisations to dynamically detect unique temporal signatures as they are reactivated in the brain. We validate this model by simulating a series of events where sequences are uniquely identifiable by analysing phasic information, as several recent EEG/MEG studies have shown. As such, we show how one can encode and retrieve complete episodic memories where the quality of such memories is modulated by the following: alpha gate keepers to content representation; binding limitations that induce a blink in temporal perception; and nested oscillations that provide preferential learning phases in order to temporally sequence events.

Memory retention in pyramidal neurons: a unified model of energy-based homo and heterosynaptic plasticity with homeostasis

The brain can learn new tasks without forgetting old ones. This memory retention is closely associated with the long-term stability of synaptic strength. To understand the capacity of pyramidal neurons to preserve memory under different tasks, we established a plasticity model based on the postsynaptic membrane energy state, in which the change in synaptic strength depends on the difference between the energy state after stimulation and the resting energy state. If the post-stimulation energy state is higher than the resting energy state, then synaptic depression occurs. On the contrary, the synapse is strengthened. Our model unifies homo- and heterosynaptic plasticity and can reproduce synaptic plasticity observed in multiple experiments, such as spike-timing-dependent plasticity, and cooperative plasticity with few and common parameters. Based on the proposed plasticity model, we conducted a simulation study on how the activation patterns of dendritic branches by different tasks affect the synaptic connection strength of pyramidal neurons. We further investigate the formation mechanism by which different tasks activate different dendritic branches. Simulation results show that compare to the classic plasticity model, the plasticity model we proposed can achieve a better spatial separation of different branches activated by different tasks in pyramidal neurons, which deepens our insight into the memory retention mechanism of brains.

Synaptic Plasticity in Cortical Inhibitory Neurons: What Mechanisms May Help to Balance Synaptic Weight Changes?

Inhibitory neurons play a fundamental role in the normal operation of neuronal networks. Diverse types of inhibitory neurons serve vital functions in cortical networks, such as balancing excitation and taming excessive activity, organizing neuronal activity in spatial and temporal patterns, and shaping response selectivity. Serving these, and a multitude of other functions effectively requires fine-tuning of inhibition, mediated by synaptic plasticity. Plasticity of inhibitory systems can be mediated by changes at inhibitory synapses and/or by changes at excitatory synapses at inhibitory neurons. In this review, we consider that latter locus: plasticity at excitatory synapses to inhibitory neurons. Despite the fact that plasticity of excitatory synaptic transmission to interneurons has been studied in much less detail than in pyramids and other excitatory cells, an abundance of forms and mechanisms of plasticity have been observed in interneurons. Specific requirements and rules for induction, while exhibiting a broad diversity, could correlate with distinct sources of excitatory inputs and distinct types of inhibitory neurons. One common requirement for the induction of plasticity is the rise of intracellular calcium, which could be mediated by a variety of ligand-gated, voltage-dependent, and intrinsic mechanisms. The majority of the investigated forms of plasticity can be classified as Hebbian-type associative plasticity. Hebbian-type learning rules mediate adaptive changes of synaptic transmission. However, these rules also introduce intrinsic positive feedback on synaptic weight changes, making plastic synapses and learning networks prone to runaway dynamics. Because real inhibitory neurons do not express runaway dynamics, additional plasticity mechanisms that counteract imbalances introduced by Hebbian-type rules must exist. We argue that weight-dependent heterosynaptic plasticity has a number of characteristics that make it an ideal candidate mechanism to achieve homeostatic regulation of synaptic weight changes at excitatory synapses to inhibitory neurons.

Distinct Heterosynaptic Plasticity in Fast Spiking and Non-Fast-Spiking Inhibitory Neurons in Rat Visual Cortex

Inhibition in neuronal networks of the neocortex serves a multitude of functions, such as balancing excitation and structuring neuronal activity in space and time. Plasticity of inhibition is mediated by changes at both inhibitory synapses, as well as excitatory synapses on inhibitory neurons. Using slices from visual cortex of young male rats, we describe a novel form of plasticity of excitatory synapses on inhibitory neurons, weight-dependent heterosynaptic plasticity. Recordings from connected pyramid-to-interneuron pairs confirm that postsynaptic activity alone can induce long-term changes at synapses that were not presynaptically active during the induction, i.e., heterosynaptic plasticity. Moreover, heterosynaptic changes can accompany homosynaptic plasticity induced in inhibitory neurons by conventional spike-timing-dependent plasticity protocols. In both fast-spiking (FS) and non-FS neurons, heterosynaptic changes were weight-dependent, because they correlated with initial paired-pulse ratio (PPR), indicative of initial strength of a synapse. Synapses with initially high PPR, indicative of low release probability ("weak" synapses), had the tendency to be potentiated, while synapses with low initial PPR ("strong" synapses) tended to depress or did not change. Interestingly, the net outcome of heterosynaptic changes was different in FS and non-FS neurons. FS neurons expressed balanced changes, with gross average (n = 142) not different from control. Non-FS neurons (n = 66) exhibited net potentiation. This difference could be because of higher initial PPR in the non-FS neurons. We propose that weight-dependent heterosynaptic plasticity may counteract runaway dynamics of excitatory inputs imposed by Hebbian-type learning rules and contribute to fine-tuning of distinct aspects of inhibitory function mediated by FS and non-FS neurons in neocortical networks.SIGNIFICANCE STATEMENT Dynamic balance of excitation and inhibition is fundamental for operation of neuronal networks. Fine-tuning of such balance requires synaptic plasticity. Knowledge about diverse forms of plasticity operating in excitatory and inhibitory neurons is necessary for understanding normal function and causes of dysfunction of the nervous system. Here we show that excitatory inputs to major archetypal classes of neocortical inhibitory neurons, fast-spiking (FS) and non-fast-spiking (non-FS), express a novel type of plasticity, weight-dependent heterosynaptic plasticity, which accompanies the induction of Hebbian-type changes. This novel form of plasticity may counteract runaway dynamics at excitatory synapses to inhibitory neurons imposed by Hebbian-type learning rules and contribute to fine-tuning of diverse aspects of inhibitory function mediated by FS and non-FS neurons in neocortical networks.

AMPA receptor trafficking and its role in heterosynaptic plasticity

Historically, long-term potentiation (LTP) and long-term depression (LTD), the best-characterized forms of long-term synaptic plasticity, are viewed as experience-dependent and input-specific processes. However, cumulative experimental and theoretical data have demonstrated that LTP and LTD can promote compensatory alterations in non-stimulated synapses. In this work, we have developed a computational model of a tridimensional spiny dendritic segment to investigate the role of AMPA receptor (AMPAR) trafficking during synaptic plasticity at specific synapses and its consequences for the populations of AMPAR at nearby synapses. Our results demonstrated that the mechanisms of AMPAR trafficking involved with LTP and LTD can promote heterosynaptic plasticity at non-stimulated synapses. These alterations are compensatory and arise from molecular competition. Moreover, the heterosynaptic changes observed in our model can modulate further activity-driven inductions of synaptic plasticity.

Hebbian plasticity requires compensatory processes on multiple timescales

We review a body of theoretical and experimental research on Hebbian and homeostatic plasticity, starting from a puzzling observation: while homeostasis of synapses found in experiments is a slow compensatory process, most mathematical models of synaptic plasticity use rapid compensatory processes (RCPs). Even worse, with the slow homeostatic plasticity reported in experiments, simulations of existing plasticity models cannot maintain network stability unless further control mechanisms are implemented. To solve this paradox, we suggest that in addition to slow forms of homeostatic plasticity there are RCPs which stabilize synaptic plasticity on short timescales. These rapid processes may include heterosynaptic depression triggered by episodes of high postsynaptic firing rate. While slower forms of homeostatic plasticity are not sufficient to stabilize Hebbian plasticity, they are important for fine-tuning neural circuits. Taken together we suggest that learning and memory rely on an intricate interplay of diverse plasticity mechanisms on different timescales which jointly ensure stability and plasticity of neural circuits.This article is part of the themed issue 'Integrating Hebbian and homeostatic plasticity'.

Multi-layer network utilizing rewarded spike time dependent plasticity to learn a foraging task

Neural networks with a single plastic layer employing reward modulated spike time dependent plasticity (STDP) are capable of learning simple foraging tasks. Here we demonstrate advanced pattern discrimination and continuous learning in a network of spiking neurons with multiple plastic layers. The network utilized both reward modulated and non-reward modulated STDP and implemented multiple mechanisms for homeostatic regulation of synaptic efficacy, including heterosynaptic plasticity, gain control, output balancing, activity normalization of rewarded STDP and hard limits on synaptic strength. We found that addition of a hidden layer of neurons employing non-rewarded STDP created neurons that responded to the specific combinations of inputs and thus performed basic classification of the input patterns. When combined with a following layer of neurons implementing rewarded STDP, the network was able to learn, despite the absence of labeled training data, discrimination between rewarding patterns and the patterns designated as punishing. Synaptic noise allowed for trial-and-error learning that helped to identify the goal-oriented strategies which were effective in task solving. The study predicts a critical set of properties of the spiking neuronal network with STDP that was sufficient to solve a complex foraging task involving pattern classification and decision making.

This study explores how intelligent behavior emerges from the basic principles known at the cellular level of biological neuronal network dynamics. Compared to the approaches used in the artificial intelligence community, we applied biologically realistic modeling of neuronal dynamics and plasticity. The building blocks of the model are spiking neurons, spike-time dependent plasticity (STDP) and homeostatic rules, known experimentally, which are shown to play a fundamental role in both keeping the network stable and capable of continous learning. Our study predicts that a combination of these principles makes possible a foraging behavior in a previously unknown environment, including pattern classification to distinct between environment shapes which are rewarded and those which are punished and decision making to select the optimal strategy to acquire the maximal number of the rewarded elements. To solve this complex task we used multi-layer neuronal processing that implemented pattern generalization by unsupervised STDP at the earlier processing step, as commonly observed in the animal and human sensory processing, followed by reinforcement learning at the later steps. In the model, the intelligent behavior emerged spontaneously due to the network organization implementing both local unsupervised plasticity and reward feedback resulting from a successful behavior in the environment.

Adenosine Shifts Plasticity Regimes between Associative and Homeostatic by Modulating Heterosynaptic Changes

Endogenous extracellular adenosine level fluctuates in an activity-dependent manner and with sleep-wake cycle, modulating synaptic transmission and short-term plasticity. Hebbian-type long-term plasticity introduces intrinsic positive feedback on synaptic weight changes, making them prone to runaway dynamics. We previously demonstrated that co-occurring, weight-dependent heterosynaptic plasticity can robustly prevent runaway dynamics. Here we show that at neocortical synapses in slices from rat visual cortex, adenosine modulates the weight dependence of heterosynaptic plasticity: blockade of adenosine A₁ receptors abolished weight dependence, while increased adenosine level strengthened it. Using model simulations, we found that the strength of weight dependence determines the ability of heterosynaptic plasticity to prevent runaway dynamics of synaptic weights imposed by Hebbian-type learning. Changing the weight dependence of heterosynaptic plasticity within an experimentally observed range gradually shifted the operating point of neurons between an unbalancing regime dominated by associative plasticity and a homeostatic regime of tightly constrained synaptic changes. Because adenosine tone is a natural correlate of activity level (activity increases adenosine tone) and brain state (elevated adenosine tone increases sleep pressure), modulation of heterosynaptic plasticity by adenosine represents an endogenous mechanism that translates changes of the brain state into a shift of the regime of synaptic plasticity and learning. We speculate that adenosine modulation may provide a mechanism for fine-tuning of plasticity and learning according to brain state and activity.SIGNIFICANCE STATEMENT Associative learning depends on brain state and is impaired when the subject is sleepy or tired. However, the link between changes of brain state and modulation of synaptic plasticity and learning remains elusive. Here we show that adenosine regulates weight dependence of heterosynaptic plasticity: adenosine strengthened weight dependence of heterosynaptic plasticity; blockade of adenosine A1 receptors abolished it. In model neurons, such changes of the weight dependence of heterosynaptic plasticity shifted their operating point between regimes dominated by associative plasticity or by synaptic homeostasis. Because adenosine tone is a natural correlate of activity level and brain state, modulation of plasticity by adenosine represents an endogenous mechanism for translation of brain state changes into a shift of the regime of synaptic plasticity and learning.