Microphysiological Systems – The Latest Research Platform Supporting Drug Discovery and Development
Part 2: Examples of microphysiological system applications in drug discoveryTranslational & Biomedical Science, Astellas Pharma Inc.
|
Welcome to the second installment of our three-part series, “The Latest Research Platform Supporting Drug Discovery and Development,” which focuses on microphysiological systems (MPS). In the first article, our colleague Kaori introduced the series by outlining the definition, key features, and growing importance of MPS, together with an explanation of relevant terminology (if you have not yet read Part 1, please take a look). In that article, she noted that one of the main hurdles to wider adoption of MPS was limited cross-company sharing of case studies. Accordingly, this second contribution presents several examples of how MPS are being used in real-world drug discovery. We hope that the information provided will serve as a useful reference for researchers who are considering introducing MPS into their own research programs.
Trends in MPS Promotion
The positioning of MPS by regulatory authorities and industry offers valuable guidance for their practical application. Before presenting individual case studies, I will first outline the key initiatives by major organizations that are advancing MPS technology.
【U.S. Food and Drug Administration（FDA）】
The FDA has been involved in the MPS field since its early days. In 2010, it co-funded a development project with the U.S. National Institutes of Health (NIH) that laid the groundwork for the first lung-on-a-chip technology [1]. In its 2017 Predictive Toxicology Roadmap, the agency identified MPS as a promising new approach for toxicity prediction [2]. More recently, the 2025 Roadmap to Reducing Animal Testing in Preclinical Safety Studies provided detailed operational guidance, including how MPS data can be incorporated into Investigational New Drug (IND) submissions and how supporting regulatory guidelines will be developed [3].
The FDA’s commitment to advancing MPS adoption is further illustrated by the Innovative Science and Technology Approaches for New Drugs (ISTAND) program, launched in 2020 [4]. This program evaluates and qualifies novel tools for specific contexts of use (CoU) in drug development [5]. Once a tool has been qualified, it may be used within its designated CoU without the need to re-evaluate the tool’s qualification for each individual drug submission, thereby shortening regulatory timelines. ISTAND was slated to become a permanent program in 2025, a transition which is expected to accelerate broader implementation.
【Innovation and Quality Microphysiological Systems (IQ MPS)】
IQ MPS is an international forum composed primarily of pharmaceutical companies which is dedicated to promoting the integration of MPS into drug development pipelines [6]. The forum organizes workshops [7-9], including those with FDA participation, and conducts surveys of industry utilization to enhance international standardization and case sharing [6]. Its publication series “IQ MPS Manuscript Series 1.0” summarizes organ-specific and topic-specific overviews of MPS characteristics and requirements [10,11]. The follow-up “IQ MPS Manuscript Series 2.0,” parts of which have already been published, addresses advanced applications such as immune systems and disease models [12-14].
【Japan Agency for Medical Research and Development (AMED)】
AMED launched the “AMED-MPS” program in 2017 to stimulate domestic development and deployment of MPS [15]. This program incorporated contributions from academia, suppliers, and pharmaceutical companies. Beginning in 2020, the Pharmaceuticals and Medical Devices Agency (PMDA) began participating as an observer in the discussion, thereby integrating regulatory perspectives [15]. Within five years, the project facilitated the development and commercialization of multiple devices. The subsequent project, AMED-MPS2, commenced in 2022, with a focus on MPS product standardization and regulatory compliance of MPS-based evaluation systems. The project aims to strengthen Japan’s competitive position in this domain [14].
Examples of MPS Implementation in Pharmaceutical Research
In the following section, I present examples of how pharmaceutical companies are leveraging MPS, including an in-house study of our own.
【Case 1: Hesperos Human-on-a-Chip®】
Hesperos’s Human-on-a-Chip® represents one of the earliest instances in which MPS-derived data were incorporated into regulatory submissions for clinical trial approval [16]. In this case, disease models of chronic inflammatory demyelinating polyneuropathy (CIDP) and multifocal motor neuropathy (MMN) were constructed to evaluate the efficacy of a Sanofi candidate compound. Induced pluripotent stem cell (iPSC)-derived motor neurons and primary Schwann cells were co-cultured with patient sera on a microelectrode array equipped with axon-guidance tunnels. The system reproduced disease-related declines in spontaneous firing frequency and conduction velocity attributable to patient autoantibodies. Treatment with the candidate compound restored these functional readouts, and the resulting data were submitted to the regulatory authority.
【Case 2: CN Bio PhysioMimix®】
CN Bio’s PhysioMimix® liver model has been validated for its utility and reproducibility through collaborative studies with the FDA [17]. The metabolic dysfunction-associated steatohepatitis (MASH) model developed on this platform closely replicates patient gene-expression profiles, and offers greater translational relevance than conventional animal models [18]. Inipharm used this model to evaluate a candidate drug for MASH and used the data to support the company’s IND submission [19,20]. In that study, hepatocytes, stellate cells, and Kupffer cells were co-cultured in a three-dimensional perfusion system while the system was under perfusion with a high-fat medium containing the test compound. After 20 days, pharmacological effects were quantified by measuring fibrosis-related markers such as type I collagen and α-SMA, as well as lipid accumulation in the medium [19]. The model is available as a ready-to-use kit, making it accessible even to researchers who are new to this platform.
【Case 3: Emulate Liver-Chip S1】
Emulate’s Liver-Chip S1 is a microfluidic model in which a porous membrane separates an upper channel seeded with primary human hepatocytes from a lower channel containing stellate cells, Kupffer cells, and liver sinusoidal endothelial cells [21]. The system enables hepatotoxicity assessment by measuring drug-induced changes in biomarkers such as albumin production and ALT release. Within the FDA’s ISTAND program, the Liver-Chip S1 was the first MPS platform to have its Step 1 Letter of Intent accepted [4]. The Step 2 Qualification Plan was subsequently accepted in 2025 [22], and the platform is now advancing to the final Full Qualification Package stage. The CoU under review is “risk assessment of drug-induced liver injury for small-molecule candidates in adults,” with the goal of informing go/no-go decisions for entry into Phase I clinical trials [22]. The platform can also apply cyclic stretch to a flexible membrane, a feature that has been applied in lung and intestinal models to mimic the dynamic functions of breathing and peristalsis, respectively [23,24].
【Case 4: Crown Bioscience Cancer Organoids】
Merus employed patient-derived tumor organoids from colorectal cancer patients to screen its bispecific antibodies. Compound efficacy was evaluated by measuring growth inhibition across organoids representing diverse genotypes and phenotypes [25,26]. To assess potential effects on normal tissue, organoids generated from the adjacent healthy colonic mucosa of the same patient were tested in the same way. Both efficacy and safety data were used to support the company’s IND submission.
【Case 5: MIMETAS OrganoPlate® 3-lane】
MIMETAS’s OrganoPlate® 3-lane features three parallel microchannels. The central channel is filled with a gel, while cells are seeded on one or both adjacent channels. Gravity-driven flow generated by a dedicated rocker supplies shear stress without the need for pumps and supports the formation of tubular structures [27]. In one study, we cultured human umbilical vein endothelial cells (HUVECs) in this platform and exposed them to candidate compounds. With edema as a potential side effect, changes in vascular permeability were evaluated. Evaluation results of the model drug are shown in Figure 1. After mitigating variability and reproducibility concerns, we rapidly measured transendothelial electrical resistance (TEER) with a dedicated reader. We then compared the extent of the TEER reduction induced by reference drugs known to cause clinical edema, as well as by other candidate compounds, with their intended pharmacological activity and used the results to support candidate selection.
We have also used the OrganoPlate platform to examine axonal outgrowth of cultured neural cells. Following a reported previously procedure [28], our in-house-generated neural progenitor cells were suspended in Matrigel and seeded into one lane while their axons were guided to extend into the central lane, thereby physically separating the somatic and axonal compartments (Fig. 2) [29]. This separation of the soma and axon regions enabled clearer visualization of axonal length, facilitating the quantitative assessment of neurotoxicity.
This case exemplifies our use of a single platform to generate multiple models that support our drug-discovery activities. The know-how we have accumulated, ranging from device handling and technical details to CoU definition, experimental design, and final data acquisition, can be carried forward to future projects. To maximize this translational value, we place particular emphasis on building models with versatility in mind.
Case studies introduced in this article are summarized in the following table.
| Model | Organ | Benefits of MPS utilization |
|---|---|---|
| Human-on-a-Chip® | Nerve | Construction of rare disease models that are difficult to evaluate in animal models, and subsequent drug assessment |
| PhysioMimix® | Liver | Efficacy evaluation including human-specific responses not reproducible in animal models |
| Liver-Chip S1 | Liver | Acceleration of regulatory submissions by leveraging the ISTAND program to qualify the test system |
| Cancer Organoids | Cancer | Dual screening of efficacy and safety using patient-derived samples to prioritize candidate antibodies |
| OrganoPlate® 3-lane | Blood vessel, nerve | Early-stage candidate selection based on toxicity evaluation of potential adverse effects |
Conclusion
In this article, I have highlighted examples of MPS implementation in drug discovery, including our own case studies. Owing to their inherent complexity, MPS often present substantial technical hurdles, and various issues must be addressed before a platform can be fully aligned with a defined CoU. Consequently, newly introduced systems are not always ready for immediate application to an ongoing project. Nevertheless, even when preliminary investigations do not progress directly to project deployment, the knowledge gained while developing MPS-based assays consistently informs and strengthens subsequent research. Through persistent, iterative exploration, we have experienced how efforts such as those described here can translate into concrete success. Equally important is the need to proactively address the factors preventing application. The third and final article in this series will focus on practical approaches to issues such as data reproducibility and assay throughput in MPS utilization. We hope that this series will deepen your understanding of MPS technology and serve as a useful reference for its implementation.
Funding Statement: The in-house case study was funded by Astellas.
（Figures and text by Maho Okubo）