The ins and outs of inhibitory synaptic plasticity: Neuron types, molecular mechanisms and functional roles

Chaotic recurrent neural networks for brain modelling: A review

Even in the absence of external stimuli, the brain is spontaneously active. Indeed, most cortical activity is internally generated by recurrence. Both theoretical and experimental studies suggest that chaotic dynamics characterize this spontaneous activity. While the precise function of brain chaotic activity is still puzzling, we know that chaos confers many advantages. From a computational perspective, chaos enhances the complexity of network dynamics. From a behavioural point of view, chaotic activity could generate the variability required for exploration. Furthermore, information storage and transfer are maximized at the critical border between order and chaos. Despite these benefits, many computational brain models avoid incorporating spontaneous chaotic activity due to the challenges it poses for learning algorithms. In recent years, however, multiple approaches have been proposed to overcome this limitation. As a result, many different algorithms have been developed, initially within the reservoir computing paradigm. Over time, the field has evolved to increase the biological plausibility and performance of the algorithms, sometimes going beyond the reservoir computing framework. In this review article, we examine the computational benefits of chaos and the unique properties of chaotic recurrent neural networks, with a particular focus on those typically utilized in reservoir computing. We also provide a detailed analysis of the algorithms designed to train chaotic RNNs, tracing their historical evolution and highlighting key milestones in their development. Finally, we explore the applications and limitations of chaotic RNNs for brain modelling, consider their potential broader impacts beyond neuroscience, and outline promising directions for future research.

Neuronal traveling waves form preferred pathways using synaptic plasticity

Traveling waves of neuronal spiking activity are commonly observed across the brain, but their intrinsic function is still a matter of investigation. Experiments suggest that they may be valuable in the consolidation of memory or learning, indicating that consideration of traveling waves in the presence of plasticity might be important. A possible outcome of this consideration is that the synaptic pathways, necessary for the propagation of these waves, will be modified by the waves themselves. This will create a feedback loop where both the traveling waves and the strengths of the available synaptic pathways will change. To computationally investigate this, we model a sheet of cortical tissue by considering a quasi two-dimensional network of model neurons locally connected with plastic synaptic weights using Spike-Timing Dependent Plasticity (STDP). By using different stimulation conditions (central, stochastic, and alternating stimulation), we demonstrate that starting from a random network, traveling waves with STDP will form and strengthen propagation pathways. With progressive formation of traveling waves, we observe increases in synaptic weight along the direction of wave propagation, increases in propagation speed when pathways are strengthened over time, and an increase in the local order of synaptic weights. We also present evidence that the interaction between traveling waves and plasticity can serve as a mechanism of network-wide competition between available pathways. With an improved understanding of the interactions between traveling waves and synaptic plasticity, we can approach a fuller understanding of mechanisms of learning, computation, and processing within the brain.

Neural Mass Modeling in the Cortical Motor Area and the Mechanism of Alpha Rhythm Changes

Investigating the physiological mechanisms in the motor cortex during rehabilitation exercises is crucial for assessing stroke patients' progress. This study developed a single-channel Jansen neural mass model to explore the relationship between model parameters and motor cortex mechanisms. Firstly, EEG signals were recorded from 11 healthy participants under 20%, 40%, and 60% maximum voluntary contraction, and alpha rhythm power spectral density characteristics were extracted using the Welch power spectrum method. Furthermore, a single-channel neural mass model was constructed to analyze the impact of parameter variations on the average power of simulated signals. Finally, model parameters were adjusted to achieve feature fitting between the simulated signals and the average power of the alpha rhythm. Results showed that alpha rhythm average power in the contralateral cortical regions increased with higher grip force levels. Similarly, the power of the simulated signals also increased with specific parameter (J, Ge, and Gi) increases, closely approximating the measured EEG signal changes. The findings suggest that increasing grip force activates more motor neurons in the motor cortex and raises their firing rate. Neural mass modeling provides a computational neuroscience approach to understanding the dynamic changes in alpha rhythms in the motor cortex under different grip force levels.

Postsynaptic dopamine D3 receptors selectively modulate μ-opioid receptor-expressing GABAergic inputs onto CA1 pyramidal cells in the rat ventral hippocampus

Although the actions of dopamine throughout the brain are clearly linked to motivation and cognition, the specific role(s) of dopamine in the CA1 subfield of the ventral hippocampus (vH) is unresolved. Prior preclinical studies suggest that dopamine D3 receptors (D3Rs) expressed on CA1 pyramidal cells exhibit a unique capacity to modulate mechanisms of long-term synaptic plasticity, but less is known about how interneuronal inputs modulate these cells. We hypothesized that inputs from μ-opioid receptor (MOR)-expressing inhibitory interneurons selectively modulate the activity of postsynaptic D3Rs expressed on CA1 principal cells to shape neurotransmission in the rat vH. We used the whole cell voltage-clamp technique to test this hypothesis by measuring evoked inhibitory postsynaptic currents (eIPSCs) from CA1 principal cells in vH slices or GABAA currents from acutely dissociated vH neurons. The eIPSC response recorded from CA1 neurons in vH slices was inhibited by either the MOR agonist DAMGO or the D3R agonist PD128907, but pretreatment with DAMGO occluded any further inhibition by PD128907. GABAA currents measured in acutely dissociated vH CA1 neurons were inhibited by D3R activation via PD128907, consistent with postsynaptic localization of D3 receptors. Kinetic alterations induced by the neuromodulatory agonists are consistent with selective targeting of postsynaptic D3Rs expressed on CA1 principal cells by MOR-expressing GABAergic inputs. Our findings suggest that postsynaptic D3R-mediated modulation of MOR-expressing GABAergic inputs is a site at which dopaminergic and opioidergic activity may contribute to disinhibition of vH excitatory neurotransmission and, thus, influence critical physiological processes such as synaptic plasticity and network oscillations.NEW & NOTEWORTHY We report that the activity of an inhibitory synapse on CA1 pyramidal cells in the rat ventral hippocampus is shaped by heterogeneous neuromodulators. Specifically, postsynaptic dopamine D3 receptors on ventral hippocampal CA1 pyramidal neurons are selectively targeted by an inhibitory input from µ-opioid receptor-expressing GABAergic terminals.

Learning-induced remodelling of inhibitory synapses in the motor cortex

Robust structural and functional plasticity occurs at excitatory synapses in the motor cortex in response to learning. It is well established that local spinogenesis and the subsequent maintenance of newly formed spines are crucial for motor learning. However, despite local synaptic inhibition being essential for shaping excitatory synaptic input, less is known about the structural rearrangement of inhibitory synapses following learning. In this study, we co-expressed the structural marker tdTomato and a mEmerald-tagged intrabody against gephyrin to visualize inhibitory synapses in layer 2/3 cortical neurons of wild-type CD1 mice. We found that a 1-day accelerated rotarod paradigm induced robust motor learning in male and female adult CD1 mice. Histological analyses revealed a significant increase in the surface area of gephyrin puncta in neurons within the motor cortex but not in the somatosensory cortex upon motor learning. Furthermore, this learning-induced reorganization of inhibitory synapses only occurred in dendritic shafts and not in the spines. These data suggest that learning induces experience-dependent remodelling of existing inhibitory synapses to fine-tune intrinsic plasticity and input-specific modulation of excitatory connections in the motor cortex.

Ultrastructural characterization of hippocampal inhibitory synapses under resting and stimulated conditions

The present study uses electron microscopy to document ultrastructural characteristics of hippocampal GABAergic inhibitory synapses under resting and stimulated conditions in three experimental systems. Synaptic profiles were sampled from stratum pyramidale and radiatum of the CA1 region from (1) perfusion fixed mouse brains, (2) immersion fixed rat organotypic slice cultures, and from (3) rat dissociated hippocampal cultures of mixed cell types. Synapses were stimulated in the brain by a 5 min delay in perfusion fixation to trigger an ischemia-like excitatory condition, and by treating the two culture systems with 90 mM high K+ for 2-3 min to depolarize the neurons. Upon such stimulation conditions, the presynaptic terminals of the inhibitory synapses exhibited similar structural changes to those seen in glutamatergic excitatory synapses, with depletion of synaptic vesicles, increase of clathrin-coated vesicles and appearance of synaptic spinules. However, in contrast to excitatory synapses, no structural differences were detected in the postsynaptic compartment of the inhibitory synapses upon stimulation. There were no changes in the appearance of material associated with the postsynaptic membrane or the length and curvature of the membrane. Also no change was detected in the labeling density of gephyrin, a GABAergic synaptic marker, lining the postsynaptic membrane. Furthermore, virtually all inhibitory synaptic clefts remained rigidly apposed, unlike in the case of excitatory synapses where ~ 20-30% of cleft edges were open upon stimulation, presumably to facilitate the clearance of neurotransmitters from the cleft. The fact that no open clefts were induced in inhibitory synapses upon stimulation suggests that inhibitory input may not need to be toned down under these conditions. On the other hand, similar to excitatory synapse, EGTA (a calcium chelator) induced open clefts in ~ 18% of inhibitory synaptic cleft edges, presumably disrupting similar calcium-dependent trans-synaptic bridges in both types of synapses.

Extracellular ATP inhibits excitatory synaptic input on parvalbumin positive interneurons and attenuates gamma oscillations via P2X4 receptors

BACKGROUND AND PURPOSE

P2X4 receptors (P2X4R) are ligand gated cation channels that are activated by extracellular ATP released by neurons and glia. The receptors are widely expressed in the brain and have fractional calcium currents comparable with NMDA receptors. Although P2X4Rs have been reported to modulate synaptic transmission and plasticity, their involvement in shaping neuronal network activity remains to be elucidated.

EXPERIMENTAL APPROACH

We investigated the effects of P2X receptors at network and synaptic level using local field potential electrophysiology, whole cell patch clamp recordings and calcium imaging in fast spiking parvalbumin positive interneurons (PVINs) in rat and mouse hippocampal slices. The stable ATP analogue ATPγS, selective antagonists and P2X4R knockout mice were used.

KEY RESULTS

The P2XR agonist ATPγS reversibly decreased the power of gamma oscillations. This inhibition could be antagonized by the selective P2X4R antagonist PSB-12062 and was not observed in P2X4-/- mice. The phasic excitatory inputs of CA3 PVINs were one of the main regulators of the gamma power. Associational fibre compound excitatory postsynaptic currents (cEPSCs) in CA3 PVINs were inhibited by P2X4R activation. This effect was reversible, dependent on intracellular calcium and dynamin-dependent internalization of AMPA receptors.

CONCLUSIONS AND IMPLICATIONS

The results indicate that P2X4Rs are an important source of dendritic calcium in CA3 PVINs, thereby regulating excitatory synaptic inputs onto the cells and presumably the state of gamma oscillations in the hippocampus. P2X4Rs represent an effective target to modulate hippocampal network activity in pathophysiological conditions such as Alzheimer's disease and schizophrenia.

Hippocampal GABAergic interneurons and memory

One of the most captivating questions in neuroscience revolves around the brain's ability to efficiently and durably capture and store information. It must process continuous input from sensory organs while also encoding memories that can persist throughout a lifetime. What are the cellular-, subcellular-, and network-level mechanisms that underlie this remarkable capacity for long-term information storage? Furthermore, what contributions do distinct types of GABAergic interneurons make to this process? As the hippocampus plays a pivotal role in memory, our review focuses on three aspects: (1) delineation of hippocampal interneuron types and their connectivity, (2) interneuron plasticity, and (3) activity patterns of interneurons during memory-related rhythms, including the role of long-range interneurons and disinhibition. We explore how these three elements, together showcasing the remarkable diversity of inhibitory circuits, shape the processing of memories in the hippocampus.

Model of ligand-triggered information transmission in G-protein coupled receptor complexes

We present a model for the effects of ligands on information transmission in G-Protein Coupled Receptor (GPCR) complexes. The model is built ab initio entirely on principles of statistical mechanics and tenets of information transmission theory and was validated in part using agonist-induced effector activity and signaling bias for the angiotensin- and adrenergic-mediated signaling pathways, with in vitro observations of phosphorylation sites on the C tail of the GPCR complex, and single-cell information-transmission experiments. The model extends traditional kinetic models that form the basis for many existing models of GPCR signaling. It is based on maximizing the rates of entropy production and information transmission through the GPCR complex. The model predicts that (1) phosphatase-catalyzed reactions, as opposed to kinase-catalyzed reactions, on the C-tail and internal loops of the GPCR are responsible for controlling the signaling activity, (2) signaling favors the statistical balance of the number of switches in the ON state and the number in the OFF state, and (3) biased-signaling response depends discontinuously on ligand concentration.

The plasticitome of cortical interneurons

Hebb postulated that, to store information in the brain, assemblies of excitatory neurons coding for a percept are bound together via associative long-term synaptic plasticity. In this view, it is unclear what role, if any, is carried out by inhibitory interneurons. Indeed, some have argued that inhibitory interneurons are not plastic. Yet numerous recent studies have demonstrated that, similar to excitatory neurons, inhibitory interneurons also undergo long-term plasticity. Here, we discuss the many diverse forms of long-term plasticity that are found at inputs to and outputs from several types of cortical inhibitory interneuron, including their plasticity of intrinsic excitability and their homeostatic plasticity. We explain key plasticity terminology, highlight key interneuron plasticity mechanisms, extract overarching principles and point out implications for healthy brain functionality as well as for neuropathology. We introduce the concept of the plasticitome - the synaptic plasticity counterpart to the genome or the connectome - as well as nomenclature and definitions for dealing with this rich diversity of plasticity. We argue that the great diversity of interneuron plasticity rules is best understood at the circuit level, for example as a way of elucidating how the credit-assignment problem is solved in deep biological neural networks.

S-Palmitoylation of Synaptic Proteins in Neuronal Plasticity in Normal and Pathological Brains

Protein lipidation is a common post-translational modification of proteins that plays an important role in human physiology and pathology. One form of protein lipidation, S-palmitoylation, involves the addition of a 16-carbon fatty acid (palmitate) onto proteins. This reversible modification may affect the regulation of protein trafficking and stability in membranes. From multiple recent experimental studies, a picture emerges whereby protein S-palmitoylation is a ubiquitous yet discrete molecular switch enabling the expansion of protein functions and subcellular localization in minutes to hours. Neural tissue is particularly rich in proteins that are regulated by S-palmitoylation. A surge of novel methods of detection of protein lipidation at high resolution allowed us to get better insights into the roles of protein palmitoylation in brain physiology and pathophysiology. In this review, we specifically discuss experimental work devoted to understanding the impact of protein palmitoylation on functional changes in the excitatory and inhibitory synapses associated with neuronal activity and neuronal plasticity. The accumulated evidence also implies a crucial role of S-palmitoylation in learning and memory, and brain disorders associated with impaired cognitive functions.

Regulation of circuit organization and function through inhibitory synaptic plasticity

Diverse inhibitory neurons in the mammalian brain shape circuit connectivity and dynamics through mechanisms of synaptic plasticity. Inhibitory plasticity can establish excitation/inhibition (E/I) balance, control neuronal firing, and affect local calcium concentration, hence regulating neuronal activity at the network, single neuron, and dendritic level. Computational models can synthesize multiple experimental results and provide insight into how inhibitory plasticity controls circuit dynamics and sculpts connectivity by identifying phenomenological learning rules amenable to mathematical analysis. We highlight recent studies on the role of inhibitory plasticity in modulating excitatory plasticity, forming structured networks underlying memory formation and recall, and implementing adaptive phenomena and novelty detection. We conclude with experimental and modeling progress on the role of interneuron-specific plasticity in circuit computation and context-dependent learning.

Evaluating the extent to which homeostatic plasticity learns to compute prediction errors in unstructured neuronal networks

The brain is believed to operate in part by making predictions about sensory stimuli and encoding deviations from these predictions in the activity of "prediction error neurons." This principle defines the widely influential theory of predictive coding. The precise circuitry and plasticity mechanisms through which animals learn to compute and update their predictions are unknown. Homeostatic inhibitory synaptic plasticity is a promising mechanism for training neuronal networks to perform predictive coding. Homeostatic plasticity causes neurons to maintain a steady, baseline firing rate in response to inputs that closely match the inputs on which a network was trained, but firing rates can deviate away from this baseline in response to stimuli that are mismatched from training. We combine computer simulations and mathematical analysis systematically to test the extent to which randomly connected, unstructured networks compute prediction errors after training with homeostatic inhibitory synaptic plasticity. We find that homeostatic plasticity alone is sufficient for computing prediction errors for trivial time-constant stimuli, but not for more realistic time-varying stimuli. We use a mean-field theory of plastic networks to explain our findings and characterize the assumptions under which they apply.

Inhibitory postsynaptic density from the lens of phase separation

To faithfully transmit and decode signals released from presynaptic termini, postsynaptic compartments of neuronal synapses deploy hundreds of various proteins. In addition to distinct sets of proteins, excitatory and inhibitory postsynaptic apparatuses display very different organization features and regulatory properties. Decades of extensive studies have generated a wealth of knowledge on the molecular composition, assembly architecture and activity-dependent regulatory mechanisms of excitatory postsynaptic compartments. In comparison, our understanding of the inhibitory postsynaptic apparatus trails behind. Recent studies have demonstrated that phase separation is a new paradigm underlying the formation and plasticity of both excitatory and inhibitory postsynaptic molecular assemblies. In this review, we discuss molecular composition, organizational and regulatory features of inhibitory postsynaptic densities through the lens of the phase separation concept and in comparison with the excitatory postsynaptic densities.

Extracellular Metalloproteinases in the Plasticity of Excitatory and Inhibitory Synapses

Long-term synaptic plasticity is shaped by the controlled reorganization of the synaptic proteome. A key component of this process is local proteolysis performed by the family of extracellular matrix metalloproteinases (MMPs). In recent years, considerable progress was achieved in identifying extracellular proteases involved in neuroplasticity phenomena and their protein substrates. Perisynaptic metalloproteinases regulate plastic changes at synapses through the processing of extracellular and membrane proteins. MMP9 was found to play a crucial role in excitatory synapses by controlling the NMDA-dependent LTP component. In addition, MMP3 regulates the L-type calcium channel-dependent form of LTP as well as the plasticity of neuronal excitability. Both MMP9 and MMP3 were implicated in memory and learning. Moreover, altered expression or mutations of different MMPs are associated with learning deficits and psychiatric disorders, including schizophrenia, addiction, or stress response. Contrary to excitatory drive, the investigation into the role of extracellular proteolysis in inhibitory synapses is only just beginning. Herein, we review the principal mechanisms of MMP involvement in the plasticity of excitatory transmission and the recently discovered role of proteolysis in inhibitory synapses. We discuss how different matrix metalloproteinases shape dynamics and turnover of synaptic adhesome and signal transduction pathways in neurons. Finally, we discuss future challenges in exploring synapse- and plasticity-specific functions of different metalloproteinases.

Plasticity in the hippocampal formation of shorebirds during the wintering period: Stereological analysis of parvalbumin neurons in Actitis macularius

The number of parvalbumin neurons can be modified by social, multisensory, and cognitive stimuli in both mammals and birds, but nothing is known about their plasticity in long-distance migratory shorebirds. Here, in the spotted sandpiper (Actitis macularius), we investigated the plasticity of parvalbumin neurons of two brain areas during this species' wintering period at a lower latitude. We compared individuals in a nonmigratory rest period (November-January) and premigration (May-July) period. We used parvalbumin as a marker for counting a subpopulation of inhibitory neurons in the hippocampal formation (HF), with the magnocellular nucleus of the tectal isthmus (IMC) as a control area. Because the HF is involved in learning and memory and social interaction and the IMC is essential for control of head, neck, and eye movements, we hypothesized that parvalbumin neurons would increase in the HF and remain unchanged in the IMC. We used an optical fractionator to estimate cell numbers. Compared with the nonmigratory rest birds, parvalbumin neuron count estimates in the premigration birds increased significantly in the HF but remained unchanged in IMC. We suggest that the greater number of parvalbuminergic neurons in the HF of A. macularius in the premigration period represents adaptive circuitry changes involved in the migration back to reproductive niches in the northern hemisphere.

The postnatal GABA shift: A developmental perspective

GABA is the major inhibitory neurotransmitter that counterbalances excitation in the mature brain. The inhibitory action of GABA relies on the inflow of chloride ions (Cl^-), which hyperpolarizes the neuron. In early development, GABA signaling induces outward Cl^- currents and is depolarizing. The postnatal shift from depolarizing to hyperpolarizing GABA is a pivotal event in brain development and its timing affects brain function throughout life. Altered timing of the postnatal GABA shift is associated with several neurodevelopmental disorders. Here, we argue that the postnatal shift from depolarizing to hyperpolarizing GABA represents the final shift in a sequence of GABA shifts, regulating proliferation, migration, differentiation, and finally plasticity of developing neurons. Each developmental GABA shift ensures that the instructive role of GABA matches the circumstances of the developing network. Sensory input may be a crucial factor in determining proper timing of the postnatal GABA shift. A developmental perspective is necessary to interpret the full consequences of a mismatch between connectivity, activity and GABA signaling during brain development.