Lab-on-Chip Microsystems for Ex Vivo Network of Neurons Studies: A Review

Hongyong Zhang, Guoguang Rong, Sumin Bian, Mohamad SawanFrontiers in Bioengineering and Biotechnology 2022 · DOI

TL;DR

This review synthesizes the latest breakthroughs in Lab-on-Chip (LoC) microsystems designed to study the complex networks of neurons (NoN) ex vivo. By combining advanced microfluidics with high-density 3D microelectrode arrays, researchers are moving closer to mimicking the human brain’s architecture to better understand and treat neurological disorders like Alzheimer’s and epilepsy.

What Problem Does This Paper Address?

Neurological disorders are a global health crisis, yet effective therapies remain elusive. The primary bottleneck is the sheer complexity of the human brain. While in vivo (in-living) studies are ethically and technically challenging, traditional in vitro (in-dish) models are often too simplistic—typically flat, 2D cultures that fail to capture the 3D connectivity and microenvironments of real brain tissue.

Standard diagnostic tools like EEGs provide a "macro" view but lack the resolution to see what individual neurons are doing. This paper explores how LoC technology bridges this gap, providing a high-resolution, controllable platform to study "networks of neurons" outside the body while maintaining biological relevance.

Methods

The paper categorizes the "Nerve-on-Chip" toolkit into three pillars:

  1. Advanced Materials: The shift from rigid silicon to flexible, biocompatible materials like PDMS and hydrogels. It also highlights the use of graphene and PEDOT:PSS for electrodes to improve transparency (allowing for simultaneous imaging) and electrical sensitivity.
  2. Microfluidic Systems: These are used for "cell manipulation"—precisely trapping single cells using dielectrophoresis (DEP) or electroosmosis—and for creating precise chemical or oxygen gradients to mimic conditions like a stroke.
  3. Microelectrode Arrays (MEAs): The evolution from 2D planar electrodes to 3D nanostructures (nanowires, nanopillars, and "mushroom-shaped" electrodes). These 3D structures can penetrate or wrap around cell membranes, allowing for intracellular recording, which is significantly more sensitive than traditional extracellular methods.

Datasets

As a review paper, this document does not introduce a new dataset. Instead, it synthesizes findings from over 140 cited primary research articles, comparing various "Nerve-on-Chip" models (e.g., lollipop-shaped nerve chips, 3D MEA chips, and organoid-on-chip systems) based on their recording methods, biomaterials, and specific applications.

Key Insights and Ideas

  • The 3D Advantage: Moving from 2D to 3D isn't just about shape; it's about signal quality. 3D electrodes (like vertical nanowires) allow researchers to record intracellular action potentials without damaging the cell, providing a much clearer picture of synaptic connectivity.
  • Microtunnels as Axon Guides: One of the most effective ways to study NoN is using microtunnels to separate the cell bodies (somas) from their long projections (axons). This allows researchers to stimulate one region and see how the signal propagates to another, effectively creating a "cerebral tract" on a chip.
  • The Rise of iPSCs: The paper emphasizes induced Pluripotent Stem Cells (iPSCs) as a game-changer. By using a patient's own cells to grow brain organoids on a chip, researchers can create personalized disease models without the ethical hurdles of animal or primary human tissue.
  • Hybrid Systems: The most advanced models now combine microfluidics (for the environment) with MEAs (for the data), allowing for long-term studies (up to a year) of how neuronal networks mature and degrade.

Key Results

The review highlights several quantitative and qualitative milestones in the field:

  • High-Density Recording: Mention of a CMOS-based MEA chip containing 65,536 electrodes, allowing for unprecedented resolution across a large network of neurons.
  • Efficiency in Cell Pairing: Microfluidic trapping techniques have reached pairing efficiencies of up to 74.2%, enabling high-throughput studies of cell-to-cell communication.
  • Disease Modeling: Successful "Parkinson’s-on-a-chip" models have been built by guiding dopaminergic projections through microtunnels to study nigrostriatal pathways.
  • Longevity: Some LoC systems have maintained viable, signaling neuronal cultures for over 12 months, essential for studying slow-progressing neurodegenerative diseases.

Limitations and Future Work

Despite the progress, the authors identify three major hurdles:

  1. Central Hypoxia: In 3D organoids, cells in the center often die because they lack "blood vessels" to deliver oxygen and nutrients. Future work must focus on vascularization-on-chip.
  2. Co-culture Complexity: Neurons don't live alone; they need glia and vascular cells. However, each cell type requires different "food" (media), making it difficult to maintain a stable, multi-cell environment on one chip.
  3. Functional Benchmarking: While we can record signals, we haven't yet trained an artificial NoN on a chip to perform a complex, "real-world" function.

The ultimate goal? Building an advanced artificial brain ex vivo that can be "trained" via chemical or electrical stimulation—a true "brain in a vat" for drug discovery and neuromorphic engineering.

Summary updated

Abstract

Increasing population is suffering from neurological disorders nowadays, with no effective therapy available to treat them. Explicit knowledge of network of neurons (NoN) in the human brain is key to understanding the pathology of neurological diseases. Research in NoN developed slower than expected due to the complexity of the human brain and the ethical considerations for <i>in vivo</i> studies. However, advances in nanomaterials and micro-/nano-microfabrication have opened up the chances for a deeper understanding of NoN <i>ex vivo</i>, one step closer to <i>in vivo</i> studies. This review therefore summarizes the latest advances in lab-on-chip microsystems for <i>ex vivo</i> NoN studies by focusing on the advanced materials, techniques, and models for <i>ex vivo</i> NoN studies. The essential methods for constructing lab-on-chip models are microfluidics and microelectrode arrays. Through combination with functional biomaterials and biocompatible materials, the microfluidics and microelectrode arrays enable the development of various models for <i>ex vivo</i> NoN studies. This review also includes the state-of-the-art brain slide and organoid-on-chip models. The end of this review discusses the previous issues and future perspectives for NoN studies.
BibTeX
@inproceedings{zhang2022lab,
  author = {Hongyong Zhang and Guoguang Rong and Sumin Bian and Mohamad Sawan},
  title = {{Lab-on-Chip Microsystems for Ex Vivo Network of Neurons Studies: A Review}},
  booktitle = {Frontiers in Bioengineering and Biotechnology},
  year = {2022},
  doi = {10.3389/fbioe.2022.841389},
}
Figure