Signal Transformations and New Timing Rules of Hippocampal CA3 to CA1 Synapses

Sandra Gattas, Aliza A. Le, Javad Karimi Abadchi, Benedict Pruess, Yanning Shen, A. Lee Swindlehurst, Michael A. Yassa, Gary LynchbioRxiv 2022 · DOI

TL;DR

This paper introduces a powerful system identification framework using Volterra series to precisely model how information is transformed at hippocampal CA3-CA1 synapses. The authors reveal previously unknown timing rules governing synaptic transmission and demonstrate that nonlinear dynamics are crucial for accurately capturing these transformations, especially in the basal dendrites.

What Problem Does This Paper Address?

Understanding how information flows through neural circuits, particularly at the synaptic level, is a fundamental challenge in neuroscience. Synaptic transmission is incredibly complex, influenced by numerous factors and past events. Previous approaches often simplified these dynamics, potentially missing crucial details. This paper aims to develop a more comprehensive and accurate mathematical model of synaptic signal transformation, specifically focusing on the CA3 to CA1 pathway in the hippocampus, to better understand its role in learning and memory.

Methods

The core methodology employed is system identification, specifically using the Volterra series expansion. This engineering approach treats the synapse as a "black box" and aims to derive an input-output transformation function. The Volterra series allows for the characterization of both linear and nonlinear system dynamics by using a series of "kernels" (h(i) for linear, h(k,m) for second-order nonlinear, and so on). The authors focused on a second-order Volterra model, which captures linear and quadratic nonlinear influences of past inputs on current outputs. They developed a "pruned" kernel approach to identify the most significant interactions and applied this to electrophysiological recordings from mouse hippocampal slices.

Datasets

The study utilized electrophysiological recordings from acute hippocampal slices of young adult male mice. Specifically, they recorded evoked field excitatory postsynaptic potentials (fEPSPs) in response to controlled stimulation of CA3 Schaffer collaterals projecting to CA1. Data was collected from 10 animals for apical dendrite recordings and 10 animals for basal dendrite recordings, with multiple slices and sessions per animal.

Key Insights and Ideas

  • Nonlinearity is King: The study strongly emphasizes that nonlinear dynamics play a critical role in synaptic signal transformation, often outweighing linear contributions, especially at higher input magnitudes.
  • Timing is Crucial: The derived transfer functions revealed previously undetected timing rules, highlighting how the precise timing of incoming spikes significantly impacts the output signal.
  • Dendritic Domain Differences: A key finding is the distinct nature of signal processing in apical versus basal dendrites. While both exhibit nonlinearities, the basal dendrites appear to require higher-order models (3rd order or beyond) for complete characterization, suggesting more complex dynamics.
  • Generalizability: The developed transfer functions demonstrated high prediction accuracy across different animals, indicating their robustness and potential for broad application.
  • A New Tool for Neuroscience: The Volterra series approach provides a sensitive and readily implementable method for evaluating the impact of pharmacological treatments and disease states on synaptic dynamics.

Key Results

  • A second-order Volterra model accurately described the CA1 apical dendritic fEPSP with >94% accuracy.
  • The nonlinear component of the transfer function was found to outweigh the linear component for input values larger than approximately 0.4 (40% of the input that induces a submaximal response).
  • The basal dendrites showed a substantially different transfer function, requiring a 3rd order system or higher for complete characterization, unlike the apical dendrites which were well-described by a 2nd order system.
  • External prediction accuracy across new animals was high, averaging 94.73% for apical dendrites and 93.56% for basal dendrites.
  • The study identified specific timing rules: rapid successive pulses (2-11 ms ISI) led to depressed responses in apical dendrites, while longer intervals (20-59 ms ISI) resulted in facilitative effects.

Limitations and Future Work

  • System Memory: The study's Volterra models had a limited memory of 60 msec. Some prediction errors were attributed to CA1 system memory extending beyond this window, suggesting that longer memory kernels might be necessary.
  • Basal Dendrite Complexity: The basal dendrites clearly require higher-order Volterra models (3rd order or beyond) for accurate characterization, indicating that the current 2nd order model is insufficient for a complete description.
  • In Vitro vs. In Vivo: The study was conducted using in vitro slice preparations. While the findings provide a strong foundation, future work should aim to validate these rules and models in vivo, where more complex network interactions occur.
  • EPSP to Spike Conversion: The current model focuses on EPSP transformation. Future research should extend this framework to predict spike patterns, which is crucial for understanding information coding in neural circuits.
  • Inter-animal Variability: While the models generalized well, some variability in individual animal estimates (particularly for the linear kernel h(i) in apical dendrites) was observed, suggesting that further investigation into the sources of this variability is warranted.

Summary updated

Abstract

Abstract The synapse is the fundamental unit of communication in the nervous system. Determining how information is transferred across the synaptic interface is one of the most complex endeavors in neuroscience, owing to the large number of contributing factors and events. An approach to solving this problem involves collapsing across these complexities to derive concise mathematical formulas that fully capture the governing dynamics of synaptic transmission. We investigated the feasibility of deriving such a formula – an input-output transformation function for the CA3 to CA1 node of the hippocampus -- using the Volterra expansion technique for nonlinear system identification. The entirety of the field EPSP in the apical dendrites of mouse brain slices was described with >94% accuracy by a 2 nd order equation that captured the linear and nonlinear influence of past inputs on current outputs. This function generalized to cases not included in its derivation and uncovered previously undetected timing rules. The basal dendrites expressed a substantially different transfer function and evidence was obtained that, unlike the apical system, a 3 rd order system or higher will be needed for complete characterization. Collectively, these results describe a readily implemented and unusually sensitive means for evaluating the effects of pharmacological treatments and disease related conditions on synaptic dynamics. At scale, the approach will also provide information needed for the construction of biologically realistic models of brain networks.
BibTeX
@misc{gattas2022signal,
  author = {Sandra Gattas and Aliza A. Le and Javad Karimi Abadchi and Benedict Pruess and Yanning Shen and A. Lee Swindlehurst and Michael A. Yassa and Gary Lynch},
  title = {{Signal Transformations and New Timing Rules of Hippocampal CA3 to CA1 Synapses}},
  year = {2022},
  doi = {10.1101/2022.05.26.493588},
  howpublished = {bioRxiv},
}
Figure