User interviews《RORZE Lifescience, Inc.》Acquiring huge amounts of data and optimizing cell culture conditions.
“The productivity of skin sheets for treating vitiligo has almost tripled, and the contamination rate has decreased by five digits.”

Originally, I personally specialized in mechanical design and equipment development, and was involved in the development of semiconductor manufacturing equipment. The opportunity to get involved in the life science market was the ‘protein crystallization robot system’ that was commissioned in 2002. After that, I created a screening system for researching drug discovery. Based on those technologies, in March 2011, we participated in the development of the automation system Tissue Factory (T-Factory^＊)for the manufacture of cell processing products, and were mainly in charge of the transportation mechanism. Since that time, and based on this experience, we have been developing automatic cell culture equipment in the field of regenerative medicine.* One of the Japanese Cabinet Office's cutting-edge research and development support programs: “the Funding Program for World-Leading Innovative R&D on Science and Technology” (FIRST). Tissue Factory for the Creation Factory (T-Factory) provides system integration for industrializing regenerative medicine.

Indeed, there were. I had started by creating culture systems without knowing much about cells, and our equipment generally did not function effectively in the beginning. However, we persevered without cease until everything operated properly, and that attitude has led us to the present moment, I believe. T-Factory was also the beginning of our relationship with Nikon. I was very grateful that people around us, including those at Nikon, understood that we were not sure of ourselves and patiently instructed us about various aspects of what we were trying to achieve. In the other direction, they also adopted our ideas which were quite different from their conventional ones, so we were able to progress with development at a reasonable cost.

Yes. I believe the BioStudio Series started with T-Factory. Nikon’s engineering staff created the first BioStudio model for us in order to fulfill our requirements, which were that it should be able to be incorporated into equipment and decontaminated. We had worked on the development of automatic observation systems for inspecting wafer defects and detecting contaminating materials within semiconductor manufacturing equipment, but Nikon’s optical system was superb.

There were two things in particular. One was how to minimize contamination. The other was how to achieve stable manufacturing, controlling the variation in cultures using a patient’s own cells (autologous cells).

Regarding the contamination reduction, the greatest source of contaminants is people, so the most effective measure is to remove them from the process. However, complete automation is not available. Therefore, we control ventilation to prevent contaminants (particles) from reaching cells. Another concern is that no machine or equipment operates completely without issue. Considering the fact that accidents do occasionally occur, such as power outages caused by lightning, for example, redundancy is built into our system design to aid recovery.

Yes, that may happen. Particularly when using autologous cells, the growth rate up to the first passage varies widely. Sometimes, extremely rare changes that would not generally be considered the norm in chemical synthesis, such as altering the parameter time freely, are made, extending the culture time by one day because growth is insufficient, for example. We are taking care to employ a flexible approach, including software that can respond to these situations.

There were actually three main reasons. The principal factor was that only the BioStudio-mini could properly observe iPS cells. We inquired with other companies about their optical system algorithms, but when we compared their algorithms and their actual way of viewing, only Nikon’s product could effectively observe iPS cells and acquire observation subject coordinates. Since we were developing and selling defect inspection systems for semiconductor wafers, we were capable of installing components such as x-y stages, and we possessed know-how about image software and tiling functions. Therefore, we were able to complete CellShot by integrating the BioStudio-mini.
The second reason was that the optical system was extremely compact.

That’s right. The development concept for CellKeeper was to use 60cm square units with various functions, combining like toy blocks. Therefore, the BioStudio-mini with its compact optical unit was ideal for installation in a limited space. There were not any serious issues with its systemization, but if I had to raise a concern, a challenge was the effect of temperature change on cells because you need to move them outside of the incubator when capturing images while performing full-scale tiled acquisition. However, we were able to solve this problem by adding temperature and CO₂ controls as standard functions within the CellShot module, which is equipped with CellKeeper.
The third reason was the flexibility of Nikon’s engineers. Based on BioStudio, they were able to heavily modify the interface. We were able to have conversations between engineers and they proceeded to completely satisfy our requirements. Without that response, the product we have now would not have been possible. I am very grateful that we were able to develop as One Team.

That’s correct. To achieve tiling, information is exchanged between our device stage and the BioStudio-mini, and they repeatedly cooperate to perform actions, which means capturing images when culture containers move. Nikon developed this information exchange into a DLL (Dynamic Link Library). Generally, it is controlled via communication, but combining programs was faster and more stable. I imagine this kind of request might usually be rejected. However, thanks to their positive response, we were able to successfully acquire live images. Also, by integration with a controlling software we developed, we were able to control the image position with an accuracy of one micron, including axial (Z) precision. For this reason, CellShot was able to realize extremely high-precision movement during tiled acquisition.

We utilize 48 culture containers (SBS standard, 100 mm dishes) that are loaded into the unit. By capturing tiled images for every plate, it has become possible to perform time-lapse analysis for the entire area within the containers. We have now achieved the macroscopic observation of cell distribution inside the culture containers and it has become possible to observe how the seeded cells  are changing within the containers.

It’s an essential tool. The amount of tiled image data captured by CellShot is enormous. No one would want to manually plot the coordinates of cells in thousands of images, right? [laughs]

Probably so. Not only that, but because the work is being carried out by human eyes, results will vary from person to person. Even if the same worker performs the task, the result may change depending on their physical condition each day. Such data cannot be effectively used for analysis. However, if this image analysis is automated, a single index can be used as a fixed standard, delivering reliable results.

Yes, I think so. Employing CellKeeper enables stable cell culture by integrating the culture process, media exchange, and observation, while also automating image analysis . As a result, we are able to clarify how various operations affect the cell culture, such as shaking the plate vertically three times after inoculation. It has brought us some surprising results.

With cultured MSCs (1,136 strains derived from newborns and 1,146 strains derived from the elderly), we have acquired time-lapse video of changes in cell distribution within culture vessels with and without various operations over seven days using CellKeeper. Doing this, we understood that whatever actions were taken, nothing was changing macroscopically from when the cells were first introduced. We released these findings at the ISBF (The International Society for Biofabrication) online event ‘Call for Abstracts for the 2021 Showcase.’

That’s correct. For example, a cell suspension is poured into the center of the dish. When this is performed, the density of the cells is high in the center but lower toward the edges. This state doesn’t change even if you turn the dish.

They actually don’t spread out. They don’t change position. On the contrary, when the suspension is poured in from the side and turned, you might assume they would gather toward the center, but they just stay where they were poured.

That’s right.

Yes, it’s quite astonishing. By employing CellShot equipped with the BioStudio-mini, all the data is captured automatically. The reproducibility is extremely high because experiments are performed without human intervention. Indeed, it feels like we have taken the first step from the ‘Meister’ method to cell culture engineering.

We have previously established mechanical functions for 1. culture (growth), 2. media exchange, 3. observation, and 4. analysis. In the future we intend to optimize culture conditions through the unified management of these functions and by further fine-tuning the technology for each of them.
As part of our latest activities, we are engaging in development and testing with MSCs (mesenchymal stem cells) at a lab in Osaka University in order to apply a subculturing system to practical use. Following that, ReMed will conduct testing using skin cells, and we will complete the work using feedback from clinical data.
After the subculturing system has been completed, we will add “5. Subculture” to the previous four functions, and furthermore, by implementing AI, we aim to achieve true automation of our cell culture systems. We also intend to advance the technological development of pretreatment and aftertreatment.

Yes, that is the case. The automation of cell culture that we are working on still does not have a market. Therefore, we are taking on the challenge from the standpoint of creating the market and the culture for it. It’s sometimes known as the ‘searching for a blue ocean strategy’, but checking market reports is not really that useful. This is because when new markets are written about in such reports, they have already been discovered. So, in fact we need to create our own ‘blue ocean strategy’. Looking at it from this perspective, I think that what we actually possess is extremely valuable and fresh information.