3D imaging of neurons cultured in microfluidic devices

Published on February 25, 2024

The Nano-Integration Devices and Systems Laboratory at the Research Institute of Electrical Communication (RIEC), Tohoku University, specializes in brain-inspired non-von Neumann computing and the foundational technologies for related hardware research. Within this laboratory, Associate Professor Hideaki Yamamoto's research group, the Nano-Integration Neurocomputing Systems Group, combines semiconductor microfabrication, nerve cell culture, and mathematical modeling to develop new in vitro systems for bottom-up analysis of brain functions. In these in vitro systems, cultured neurons occasionally aggregate to form a 3D structure. The group introduced the MAICO^® MEMS confocal unit for 3D imaging of these aggregated neurons.

We interviewed Prof. Hideaki Yamamoto and Mr. Hakuba Murota, who is responsible for cell imaging and analysis using the MAICO MEMS confocal unit, to learn about the background of implementing the MAICO MEMS confocal unit, their experience using it, and future research prospects.

Current research
Challenges in Neuronal Imaging
Decisive factor for introducing the MAICO MEMS confocal unit
Usability of the MAICO MEMS confocal unit
Imaging examples
Prospects for research
About Multicellular Neurobiocomputing: Understanding and Advancing towards Biological Supremacy
Researcher profile

Current research

Could you tell us about your research?

Our current research focuses on developing new in vitro systems that can serve as models for the complex network of neurons that constitute animal brains. For example, while research on the heart or cancer has advanced to the point where physiological functions and pathological conditions can be reproduced using cultured cells, there are currently no effective model systems for extremely complex tissues like the brain, where we see a critical issue. For instance, when nerve cells obtained from rat cortices are cultured on a flat dish, they form a network that exhibit activity patterns different from those in living tissue. Therefore, one of our research motivations is to make these activity patterns more like to those in brain tissue. Our group has a background in electronics. Therefore, while new cell culture technologies like organoids are rapidly advancing, we want to realize the goal by using semiconductor manufacturing technologies to reproduce local wiring structures in the brain with living cells.

Specifically, using equipment in our semiconductor clean room (Figure 1), we create microfluidic devices that guide nerve cells to adhere to and extend their neurites (Figure 2). When neural cells are cultured in microfluidic devices, they form dense connections within each cell (through-hole) and then extend their neurites through microchannels connecting the cells to form neural networks. This structure - where densely connected cell populations weakly interact with each other - is said to be a characteristic of actual cerebral cortical neural circuits. By culturing neural cells using this device, we are conducting experiments to partially reproduce network structures similar to those seen in the brain.

Associate professor Hideki Yamamoto

Clean room where microfluidic devices and other devices are fabricated

Figure 1: Clean room where microfluidic devices and other devices are fabricated

Neurons cultured in a microfluidic device

Figure 2: Neurons cultured in a microfluidic device

Challenges in Neuronal Imaging

Mr. Hakuba Murota

What are the challenges of imaging neurons?

When we image cultured nerve cells, we typically use an epifluorescence microscope with a sCMOS camera attached. However, when cell culture continues for several days, region with high cell density form aggregates and develop into 3D structures. With the conventional combination of an epifluorescence microscope and a sCMOS camera, we couldn't image these structures in 3D. As a result, our group had been exploring culture methods to minimize cell aggregation. However, if the system could function well even with aggregated cells, we wanted to keep this characteristic.

This created the need for a confocal microscope to observe these 3D structures, but at the time, our laboratory didn't have a confocal microscope, nor was there one available in our facility's shared equipment. While confocal microscopes were available at other campuses, transporting live samples were cumbersome, so we wanted a confocal microscope that we could use within our research facility. We faced a dilemma, because confocal microscopes are generally expensive, making it difficult to acquire one without securing a substantial budget.

Decisive factor for introducing the MAICO MEMS confocal unit

What made you decide to introduce MAICO?

When we were facing the problem of not having a confocal microscope in our laboratory, we came across information at an exhibition about a new unit from Hamamatsu Photonics that could be attached to an existing microscope to build a confocal microscope by. This caught our interest, and in the following year, we arranged a product demonstration. With its satisfactory sensitivity and image quality, we immediately decided to adopt it.

The most decisive factor was the reasonable price. Our laboratory initially introduced the unit with a single 488 nm wavelength configuration. We remember being very grateful for the budget-friendly aspect as it cost less than 5 million yen, which was extremely reasonable. Additionally, we appreciated that it had a subunit structure allowing us to add necessary wavelengths as needed. Currently, we are operating with a two-wavelength configuration after adding a 638 nm unit. Looking ahead, we are considering adding 405 nm and 561 nm units, depending on the progress of our research.

Imaging system using MAICO

They also use our ORCA-Fusion BT for regular epifluorescence imaging.

Usability of the MAICO MEMS confocal unit

Prof. Yamamoto and Mr. Murota

How would you feel about the usability of MAICO?

We haven't extensively used confocal microscopes from other manufacturers, so we can't make direct comparisons, but we find it attractive that we can start imaging in about 15 minutes after powering on the main unit. The HCImage software used to control MAICO is also intuitive and easy to use. Additionally, the resolution simulator provided by Hamamatsu Photonics is nice, which allows us to easily check in advance the achievable resolutions.

Currently, we are using MAICO for imaging aggregated neurons and are very satisfied with the clear visualization of their 3D structures. Furthermore, even for neurons growing in 2D, MAICO provides higher-resolution compared to a standard camera. While we are currently using a sCMOS camera for calcium imaging of neurons, MAICO could potentially be used for calcium imaging of aggregates in the future, given its fast frame rate.

Imaging examples

3D images of aggregated rat cortical neurons (imaging NeuO and GCaMP6s fluorescence).

Murota et al., Advanced Materials Technologies (2024).

Data provided by: Hideaki Yamamoto, Nano-Integration Devices and Systems Laboratory for Nanoelectronics and Spintronics, Research Institute of Electrical Communication, Tohoku University

Axons and dendrites in cultured rat cortical neurons.

max projection image of neurons captured outside a microfluidic device

This is a max projection image of neurons captured outside a microfluidic device.

Data provided by: Hideaki Yamamoto, Nano-Integration Devices and Systems Laboratory for Nanoelectronics and Spintronics, Research Institute of Electrical Communication, Tohoku University

Prospects for research

Could you tell us about future research prospects?

The first goal is to increase the number of wells in the microfluidic device to construct a more complex network. Previously, we cultured neurons in a 2 × 2 well array, but now we have created a microfluidic device with a 4 × 4 well array. By increasing the number of wells and thus the number of neurons, we are trying to replicate more complex neuronal networks. Additionally, in the future, we may create 3D microfluidic devices with microchannels arranged not only in the XY direction but also in the Z direction. Although creating 3D structures with the current fabrication process remains challenging, the ability to perform 3D observations with a confocal microscope encourages us to take on such challenge.

The second goal is calcium imaging of neuronal aggregates. Until now, we have primarily performed calcium imaging of neurons cultured 2D, either on cover glasses or within microfluidic devices. However, with access to MAICO, we now have the capability to perform calcium imaging on aggregated neurons. If this works well, it could open new directions in our research, such as actively culturing neurons in aggregates.

Prof. Yamamoto

About Multicellular Neurobiocomputing: Understanding and Advancing towards Biological Supremacy

Could you tell us about the newly launched “Multicellular Neurobiocomputing: Understanding and Advancing towards Biological Supremacy” project?

This project was launched in April 2024 with support from the Grant-in-Aid for Transformative Research Areas (A). The phrase "Advancing towards Biological Supremacy" in the project title signifies the idea that systems inspired by biological mechanisms can solve specific problems with learning efficiency, energy efficiency, and environmental adaptability that are difficult to achieve with conventional computers. As I mentioned earlier, our brain is formed by neurons that are connected to create networks. Unlike transistors, or the elements of integrated circuits in computers and smartphones, neurons are inherently unstable. However, our brain performs complex information processing autonomously and with high energy efficiency. This property does not arise from a single cell and cannot be explained simply as a sum of individual elements. Instead, brain functions emerge through the sophisticated arrangement and wiring of various types of neurons, forming a multicellular network. In this project, our goal is to understand how the brain uses networks of biological elements to process infromation by combining mathematical modeling with in vivo and in vitro experiments, and translate this understanding into practical system applications. This research is expected to contribute not only to a fundamental understanding of the nervous system but also to the development of innovative computing technologies with high computational efficiency, robustness, and adaptability.

To achieve this, the project brings together researchers from diverse fields, including information science, bioengineering, biology, and electronics. We have set three key reserch domains: Multicellular Modeling (the formulation of information processing models and learning rules based on biological experiments), Multicellular Hardware (the implementation of functions in hardware and application to robotics), and Multicellular Wetware (the verification of mathematical models and learning rules using cultured cells and the artificial reconstruction of biological functions). Through these domains, we are challenging to demonstrate Biological Supremacy, aiming to harness the principles of biology in next-generation information and communication technology.

About Multicellular Neurobiocomputing: Understanding and Advancing towards Biological Supremacy

Click here for the website of the Grant-in-Aid for Transformative Research Areas (A), Multicellular Neurobiocomputing: Understanding and Advancing towards Biological Supremacy.

Researcher profile

Prof. Hideaki Yamamoto

Hideaki Yamamoto
Associate Professor, RIEC, Tohoku University
Associate Professor, Advanced Institute for Materials Research (AIMR), Tohoku University
Associate Professor, Department of Electrical, Information and Physics Engineering, Tohoku University
Associate Professor, Department of Electronic Engineering, Graduate School of Engineering, Tohoku University

Mar. 2009

Ph.D., Department of Nanoscience and Nanoengineering, Graduate School of Advanced Science and Engineering, Waseda University

Apr. 2009

JSPS Research Fellowship for Young Scientists (PD), Waseda University

Apr. 2010

JSPS Research Fellowship for Young Scientists (SPD), Tokyo University of Agriculture and Technology

Apr. 2013

Assistant Professor, Waseda Institute for Advanced Study (WIAS)

Apr. 2014

Assistant Professor, Frontier Research Institute for Interdisciplinary Sciences (FRIS), Tohoku University

May. 2018

Assistant Professor, AIMR, Tohoku University

Jan. 2020

Current Role

Mr. Hakuba Murota

Hakuba Murota
Ph.D., Nano-Bio Hybrid Molecular Devices Laboratory, RIEC, Tohoku University

Mar. 2022

Bachelor's Degree in Engineering, Department of Electrical, Information and Physics Engineering, Tohoku University

Mar. 2024

Master's Degree in Engineering, Department of Electronic Engineering, Graduate School of Engineering, Tohoku University

Apr. 2024

Doctoral course, Department of Electronic Engineering, Graduate School of Engineering, Tohoku University

Click here for the website of Nano-Integration Devices and Systems Laboratory for Nanoelectronics and Spintronics, Research Institute of Electrical Communication, Tohoku University.

Other case studies

3D live cell imaging analysis of epidermal keratinocytes

The Laboratory for Evolutionary Cell Biology of Skin, Cosmetics Course, School of Bioscience and Biotechnology, Tokyo University of Technology, is researching epidermal barrier formation mechanisms. To elucidate the mechanisms of epidermal barrier formation, it is necessary to image the epidermis in three dimensions. For this purpose, they have introduced our MAICO® MEMS confocal Unit.

We interviewed Professor. Takeshi Matsui from the same laboratory about the background of introducing the MAICO MEMS confocal unit, his impressions of its use, and the prospects for future research.