Extracellular detection of neuronal coupling

Elmer Guzman, Zhuowei Cheng, Paul K. Hansma, Kenneth R. Tovar, Linda Petzold, Kenneth S. KosikScientific 2021 · DOI

TL;DR

This paper introduces a novel, non-invasive method for detecting synaptic connections between neurons using extracellular recordings from microelectrode arrays (MEAs). By analyzing the propagation of extracellular action potentials (eAPs) along single axons, the researchers can reliably identify and characterize neuronal coupling, offering a more robust alternative to traditional spike-sorting techniques.

What Problem Does This Paper Address?

Understanding how neurons communicate is fundamental to neuroscience. Traditionally, this involves recording electrical activity from neurons, often using techniques like patch-clamping or extracellular recordings. However, extracellular recordings from MEAs, while capable of capturing signals from many neurons simultaneously, face a significant challenge: distinguishing the signals originating from individual neurons. When neurons are cultured on MEAs, their extracellular action potentials (eAPs) can overlap, making it difficult to assign a specific spike to its source neuron. This ambiguity is particularly problematic when trying to map synaptic connections, as it requires accurately identifying pre- and post-synaptic activity. Existing spike-sorting algorithms often struggle with this, and their validation can be challenging. This paper aims to overcome this limitation by developing a method that leverages the physical propagation of eAPs along axons to unambiguously identify individual neurons and then use this identification to detect synaptic coupling.

Methods

The core of this work lies in a two-step algorithmic approach:

  1. Propagation Signal Detection: The method first identifies "propagation signals." These are eAPs that are detected sequentially across multiple electrodes in a consistent spatial and temporal pattern, indicative of an action potential traveling down a single axon. By analyzing the latency and consistency of these signals across electrode pairs, the algorithm can reliably label eAPs originating from individual neurons. This approach bypasses the need for complex spike sorting.
  2. Short Latency Coupling Detection: Once individual neurons are identified by their propagation signals, the algorithm uses these signals as temporal "clocks." It then looks for clusters of eAPs occurring at other electrodes within a short, defined time window (0.5-10 ms) after the propagation signal spike. These short-latency events are hypothesized to be postsynaptic responses.

To validate these detected couplings, the researchers employed several strategies:

  • Amplitude Distribution Analysis: They compared the amplitude distribution of these "coupled" spikes to the overall amplitude distribution at the target electrode. Coupled spikes often formed a distinct subset of the overall distribution, suggesting they originated from a single, specific postsynaptic neuron.
  • Direct Stimulation: They stimulated identified presynaptic neurons and observed if the postsynaptic response recapitulated the characteristics of the spontaneously detected couplings in terms of probability and amplitude.
  • Pharmacological Manipulation: They used Cadmium (Cd²⁺) to reduce neurotransmitter release probability and observed a corresponding decrease in coupling probability, as expected for synaptic connections.
  • Temperature Sensitivity: They investigated the effect of temperature on coupling latency, finding that increased temperature led to shorter latencies, consistent with known effects on synaptic transmission and action potential propagation.

Datasets

The study utilized in vitro neuronal cultures derived from mouse hippocampal neurons plated on multi-electrode arrays (MEAs). These cultures served as the primary data source for recording neuronal activity. No specific external datasets were mentioned for training or evaluation; the algorithms were developed and validated using data generated from these in vitro preparations.

Key Insights and Ideas

  • Propagation as a Neuron Identifier: The key insight is that the physical propagation of an action potential along an axon provides a robust, non-invasive signature for identifying individual neurons, circumventing the difficulties of spike sorting.
  • eAP Propagation as a "Clock": Identified propagation signals act as precise temporal markers, allowing for the accurate timing of potential postsynaptic events.
  • Short Latency as a Synaptic Indicator: The detection of short-latency spike clusters following a propagation signal is a strong indicator of synaptic coupling, especially when combined with other validation criteria.
  • Functional Connectome Emergence: The method allows for the assembly of a "functional subset" of neuronal connectivity within the cultured network, revealing relationships that might be missed by other techniques.
  • Latent Connectivity: The Cadmium experiments suggest the existence of "latent connectivity" – connections that are not readily apparent under normal conditions but can be revealed by modulating synaptic strength or inhibitory tone.

Key Results

  • The developed algorithm successfully identified propagation signals from individual neurons with high fidelity, characterized by low inter-electrode latency and low coefficient of variation.
  • Short-latency spike clusters following propagation signals were reliably detected and validated as likely synaptic couplings.
  • Direct stimulation of presynaptic neurons recapitulated the probability and amplitude characteristics of spontaneously detected coupled spikes.
  • Reducing active synapse density with Cadmium significantly decreased coupling probability, confirming the synaptic nature of the detected connections.
  • Temperature changes affected coupling latency in a manner consistent with known physiological effects on synaptic transmission and action potential propagation.
  • The method demonstrated the ability to identify multiple presynaptic inputs to a single postsynaptic neuron and quantify their contribution.

Limitations and Future Work

  • Electrode Density: The authors acknowledge that the relatively low electrode density of MEAs means that not all neurons or connections within the culture can be captured, leading to an incomplete "functional connectome."
  • Indirect Coupling: While the data strongly suggests direct synaptic coupling, the authors concede they cannot entirely rule out indirect synaptic coupling, though the observed latencies are more consistent with direct connections.
  • Inhibitory Neurons: The study notes that Cadmium can also affect inhibitory neurons, potentially influencing the observed redistribution of connections. Further work could explore the specific roles of inhibitory connections.
  • Generalizability: While the method is presented as generalizable to mutant or iPS-derived neurons, further validation across diverse neuronal types and network structures would be beneficial.
  • Future Directions: The authors suggest that this method can be used to screen the effects of pharmacological manipulations and to characterize synaptic connectivity phenotypes in genetically modified or iPS-derived neurons, opening avenues for disease modeling and drug discovery. The ability to perform long-term monitoring without the invasiveness of patch-clamp recording is also highlighted as a significant advantage.

Summary updated

Abstract

We developed a method to non-invasively detect synaptic relationships among neurons from in vitro networks. Our method uses microelectrode arrays on which neurons are cultured and from which propagation of extracellular action potentials (eAPs) in single axons are recorded at multiple electrodes. Detecting eAP propagation bypasses ambiguity introduced by spike sorting. Our methods identify short latency spiking relationships between neurons with properties expected of synaptically coupled neurons, namely they were recapitulated by direct stimulation and were sensitive to changing the number of active synaptic sites. Our methods enabled us to assemble a functional subset of neuronal connectivity in our cultures.
BibTeX
@inproceedings{guzman2021extracellular,
  author = {Elmer Guzman and Zhuowei Cheng and Paul K. Hansma and Kenneth R. Tovar and Linda Petzold and Kenneth S. Kosik},
  title = {{Extracellular detection of neuronal coupling}},
  booktitle = {Scientific},
  year = {2021},
  doi = {10.1038/s41598-021-94282-6},
}
Figure