Novel organ-on-chip systems for predictive toxicology and drug development.

As cardiac diseases are one of the main causes of death worldwide, research on better understanding its mechanisms and relation to drugs is fundamental in reducing their burden on society. To better understand the mechanisms of heart disease, current research has focused on in vivo animal studies and in vitro cardiac models. However, animal studies may raise ethical concerns and do not fully predict human responses. Additionally, traditional in vitro studies overly simplify the heart by analyzing the activity of a single type of ion channel protein present in cardiac cells. These results are neither sensitive nor fully predictive.

In an effort to better predict the risks of adverse cardiac effects, the focus of in vitro cardiotoxicity assays is shifting away from traditional models. A wide range of devices were developed to monitor the electrical or contractile activities of cardiomyocytes in vitro. Moreover, the phenotype of cardiac cells grown on traditional petri dishes look strikingly different from in vivo cardiac cells. To tackle this, heart-on-a-chip devices are being developed to engineer the microenvironment around cells and to better simulate properties of the heart. So far, most systems developed suffer from disadvantages such as low-throughput of measurements and of fabrication, limited signal quality, and only measure either the electrical or the contractile activity of cardiac cells.

The goal of this work was to develop an integrated heart-on-a-chip platform to improve in vitro cardiac research. This work envisioned the integration of non-invasive sensors, engineered microenvironments, and scalable microfluidic systems onto a single silicon-based heart-on-a-chip platform. A major requirement for this platform was the compatibility of the different applied technology modules to each other and the ability to be mass-produced. Due to the multidisciplinary character of this project, this manuscript separately addresses the three main areas for improvements for heart-on-a-chip devices.

The first part of this work discusses novel sensors to monitor cardiac activity in vitro. For this purpose, a lens-free imaging (LFI) strategy to non-invasively monitor the deformation of cardiac cells cultured on transparent or opaque substrates was developed. By integrating a reflection-based LFI system onto high-density microelectrode arrays (MEA) chips, both the intracellular electrical activity and the contractility of cardiac cells were simultaneously monitored. This work analyzed drug-induced alterations in microscopic (cardiac deformation and excitation-contraction coupling) and macroscopic (propagation of the excitation wave) cellular phenomena. The combination of electrical and contractile sensors into 1 device is of great value for cardiotoxicity screening and drug development as it offers unique insight into adverse cardiac effects.

 The second part of the work presents two strategies to improve the relevance of in vitro cell models. In one approach, the use of PisC, a smart polymer material with photo-patternable surface properties (JSR Corporation, Tokyo, Japan), was evaluated to determine if the adsorption of proteins and cells on substrates could be influenced. By creating lines of cell-adhesive regions, cardiac cells grew into elongated, myofiber-like structures. In the second approach, nanogrooves etched directly into the silicon chip surfaces guided the growth of cells. In both approaches, cardiac cells aligned and elongated according to the direction of the pattern, inducing morphologies that have the potential to better resemble the in vivo situation.

The third part of the work tackles the limited throughput of heart-on-a-chip assays. For this purpose, a novel strategy to couple polystyrene and glass-based microfluidic cell culture systems onto a silicon-based MEA was evaluated. The microfluidic system divided the MEA surface into 16 separate chambers. Cardiac cells were successfully cultured for a week in prototype devices. The functionality of cardiac cells was confirmed by LFI contractility measurements on transparent microfluidic devices. Finally, the extracellular electrical activity of cardiac cells was recorded in a proof-of-concept platform wherein a microfluidic system is integrated onto a high-density MEA. These results showed that cardiac cultures and silicon-based sensors are compatible with microfluidic devices, which enables the upscaling of experiment throughput.

This thesis demonstrates the great potential of silicon-based heart-on-a-chip device for improving in vitro cardiac research. Future research on this heart-on-a-chip platform should focus on more relevant, human cardiac models using inducible pluripotent stem cell-based reprogramming strategies. These cell models are currently used for disease modeling purposes and would offer more relevant results as they are more predictive of the human response. Stem cell technologies could thus enable personalized heart-on-a-chip models based on a patient’s own cells. Alternatively, assays with cardiac progenitor cells isolated from the human heart could lead to a better understanding of cardiac regeneration and regenerative medicines. Ultimately, such heart-on-a-chip systems may lead to improved treatments for and reduced burden of cardiac diseases.