Abstract Patterns of waxing and waning oscillations, called spindles, are observed in multiple brain regions during sleep. Spindles are thought to be involved in memory consolidation. The origin of spindle oscillations is ongoing work, but experimental results point towards the thalamic reticular nucleus (TRN) as a likely candidate. The TRN is rich in electrical synapses, also called gap junctions, which promote synchrony in neural activity. Moreover, gap junctions undergo activity-dependent long-term plasticity. We hypothesized that gap junction plasticity can modulate spindle oscillations. We developed a computational model of gap junction plasticity in recurrent networks of TRN and thalamocortical neurons (TC). We showed that gap junction coupling can modulate the TRN-TC network synchrony and that gap junction plasticity is a plausible mechanism for the generation of sleep-spindles. Finally, our results are robust to the simulation of pharmacological manipulation of spindles, such as the administration of propofol, an anesthetic known to generate spindles in humans.
Introduction
Sleep spindles are transient bursts of 7–15 Hz neural activity that characterize non-REM sleep. These rhythms are not merely biological noise; they are fundamentally linked to the process of memory consolidation and the maintenance of sleep. While the thalamus has long been identified as the source of these oscillations, the precise mechanism that governs their "waxing and waning"—the periodic increase and decrease in amplitude—remains a subject of intense research.
Traditional models of spindle generation focus on the feedback loops between excitatory thalamocortical (TC) neurons and inhibitory neurons in the Thalamic Reticular Nucleus (TRN) mediated by chemical synapses. However, the TRN is uniquely characterized by a high density of electrical synapses, or gap junctions. Unlike chemical synapses, gap junctions provide direct electrical continuity between cells and are known to undergo activity-dependent long-term plasticity. This study explores whether the plasticity of these electrical connections could be the primary driver of the spindle rhythm.
Computational Modeling of the Thalamic Network
To investigate this hypothesis, we developed a computational network model representing the interaction between TRN and TC populations.
Neuron Models
- TRN Neurons: These were modeled using Izhikevich equations, which allow for the simulation of the complex firing patterns observed in the TRN, specifically the ability to switch between tonic spiking and high-frequency bursting.
- TC Neurons: These were represented as Leaky Integrate-and-Fire (LIF) neurons, capturing the basic integrative properties of excitatory thalamic cells.
Gap Junction Plasticity Rule
The core innovation of this model is the implementation of a biophysically inspired plasticity rule for gap junctions within the TRN. This rule is bidirectional:
- Long-Term Depression (gLTD): Sustained, high-frequency bursting activity causes the electrical coupling between neurons to weaken.
- Long-Term Potentiation (gLTP): Periods of sparse, low-frequency firing allow the gap junctions to strengthen.
By simulating these dynamics using Euler’s method and analyzing the output through Fourier analysis, we were able to quantify the power and frequency of the resulting oscillations.
The Mechanism: A Loop of Synchrony and Plasticity
The simulations revealed that the network exists in a state of constant tension between two regimes: the Asynchronous Irregular (AI) state and the Synchronous Regular (SR) state. The "waxing and waning" of spindles emerges from a continuous loop between these states:
- The Waxing Phase: When the network is in a quiet, asynchronous state, the gLTP rule dominates. Gap junctions gradually strengthen, increasing the coupling between TRN neurons.
- The Spindle Burst: Once the coupling reaches a critical threshold, the TRN neurons synchronize. This synchrony triggers rhythmic bursting at 7–15 Hz, which is transmitted to the TC neurons, creating the visible "spindle."
- The Waning Phase: The intense bursting activity during the spindle triggers the gLTD rule. The gap junctions weaken, the synchrony collapses, and the network returns to the AI state, ending the spindle.
Results and Mathematical Dynamics
Spindle Characteristics and Chaos
The model successfully produced oscillations matching biological sleep spindles, with frequencies of 7–15 Hz occurring at intervals of 5–15 seconds. Interestingly, these spindles were not perfectly periodic. By calculating the Lyapunov exponent, we determined that the system behaves as a chaotic attractor. This mathematical property explains the inherent variability in spindle duration and timing observed in real-world EEG recordings.
Pharmacological Validation
To test the robustness of the model, we simulated the effects of propofol, an anesthetic known to induce spindles in humans. Propofol acts by increasing the time constant of inhibitory GABAergic synapses. In our model, this shift moved the network's fixed point into a regime where gap junction plasticity could more easily trigger oscillations, aligning our computational findings with clinical observations.
TRN Autonomy
Our results also demonstrated that the mechanism is robust even in isolated TRN networks. This supports in vitro evidence suggesting that the TRN can act as an independent pacemaker for spindle-like activity, with the TC neurons primarily serving to relay and amplify the signal to the cortex.
Discussion and Limitations
While this model provides a compelling mechanism for spindle generation, it highlights areas for further experimental inquiry. Specifically, while gap junction depression (gLTD) is well-supported by experimental data, the evidence for activity-dependent potentiation (gLTP) in the TRN is currently more circumstantial. Confirming this "up-regulation" in biological tissues is a critical next step.
Furthermore, as a "minimal substrate" model, this study does not account for the extensive feedback the thalamus receives from the cortex. Future iterations of the model will need to incorporate these top-down influences to understand how the cortex modulates the timing and intensity of spindles during different stages of sleep. Despite these limitations, the study establishes gap junction plasticity as a plausible and powerful engine for the rhythmic architecture of the sleeping brain.
Summary updated