The following article originally appeared on September 20, 2020 in Drug Discovery World here.
Dr. Kacey Ronaldson-Bouchard explains why adapting cell-based assays from suspension cells to 2D culture, 3D culture, and beyond to ‘patient in a dish’ systems is not an exercise undertaken lightly. However, the effort pays off on many levels.
Our incomplete understanding of human biology and the limited potential for human experimentation underscore the need for in vitro models of human diseases. Ideally, such a system would accurately recapitulate disease etiology and progression, predict responses to treatment, and serve as a patient- or species-specific test platform. Preclinical drug testing has traditionally relied on animal models, but cell-based assays (CBAs) are taking on more and more of the workload. In this context, assay platforms tend towards greater complexity and functionality, leading to a higher degree of similarity between in vitro and in vivo systems.
Along the continuum of physiologic relevance for cell-based assays, two-dimensional (2D) cultures are the simplest and arguably still the most widely used, followed by multi-cellular 3D spheroids and organoids. Further along we have tissue chips, the subject of this article, and human-on-chip platforms which are still in early development.
2D cultures of attachment-dependent cells are easily realised but they fail to replicate the chemical, mechanical, and physiologic factors affecting cells in living tissues. Particularly lacking are cell-cell and cell-extracellular interactions responsible for cell differentiation, proliferation, vitality, genes and protein expression, responses to stimuli, and drug metabolism.
Three-dimensional (3D) cell culture adds deep physiologic relevance to cell-based assays by recapitulating the in vivo chemical, mechanical, and biological conditions that cells experience during healthy and disease states. When cultured from patient-derived cells, spheroids and organoids incorporate the subject’s genomic context as well, including relevant efficacy and toxicology responses to existing or experimental drugs. 3D culture could, therefore, provide the link between one-size-fits-all treatments and personalised medicine.
Personalised medicine or precision medicine (PM) seeks to deliver the right drug to the right patient, in the right dose and at the right time. PM involves using a companion diagnostic – typically a biochemical assay – to stratify patients for clinical studies, or to determine their suitability for specific treatments. According to the US Food and Drug Administration, a companion diagnostic “provides information that is essential for the safe and effective use of a corresponding drug or biological product.” 3D cell culture generally, and tissue chips specifically, have the potential to fine-tune stratification to include specific safety, efficacy, and inter-organ responses. By adding multiple readouts from two or more in vitro organ systems, tissue chips can simultaneously address a further aspect of personalisation: why some patients get sick at all and some, with the same frank susceptibility, do not.
Spheroids are simple clusters of cells derived from almost any tissue, for example from tumor biopsies, embryoid bodies, hepatocytes, nervous tissue, or mammary glands. Spheroids do not require any special scaffolding or engineering as they self-assemble through cellular attraction and adhesion. Spheroids also do not replicate, which limits their manufacturability, availability for large-scale screening, and batch-to-batch consistency.
Organoids are derived from a single organ-specific stem- or progenitor cell from stomach, liver, bladder, or other organs. When cultured within an extracellular scaffold and provided with suitable culture media and growth factors, seed cells differentiate into all cell types normally found in the original tissue and self-organize into usable structures.
The terms ‘tissue chip’ and ‘organ chip’ are often used interchangeably but the differences are notable. Organ chips are microfluidic systems with controlled, dynamic microenvironments in which cultured cells emulate organ-level physiology.
Organ chips typically incorporate one or two cell types cultured in a polymer microfluidic device, which emulates one specified function of an organ. Tissue chips, the term preferred by tissue engineers, are larger and more highly engineered. Tissue chips are constructed by preparing its cellular constituents individually, combining them in their physiologic proportions, and engineering-in niche-level biomimetic forces when applicable. Tissue chips incorporate microfluidics to connect multiple in vitro organ/tissue models, whereas these functional units are independent and can also be used free-standing in organ chips (as well as in organoids). In part due to their greater complexity compared with either organ chips or organoids, tissue chips are lower-throughput test platforms that should provide increased physiological relevance.
Since they derive from patient-harvested pluripotent stem cells, tissue chips are patient-specific, and of course species-specific – a fully human model. By providing physiologically relevant readouts tissue chips support disease monitoring, drug responses, and personalised medicine.
Like other human-derived cell-based assays, tissue chips compliment well-validated animal toxicology models, but unlike advanced 3D organoids their functionality is limited by their constituent cells. Further functionality may be introduced in the form of adding additional cell types and physical or mechanical constraints typical of the tissue’s original niche, which coaxes the tissue to mature by delivering biologically relevant signals. Additional incorporation of microfluidics enables the study of two or more organ functions under physiologic conditions, including at the same dosing levels.
‘Patient-in-a-dish’ refers to chip-based organ systems interconnected for tissue-tissue communication, in which individual ‘organs’ experience the same stimulus, to the same degree, simultaneously. When designed for all characteristics relevant to disease progression or drug response, patient-in-a-dish systems facilitate the study of systemic diseases, including those that affect organs differently or at different stages of treatment or disease status. Additional factors may be built in as well, for example immune system cells that confer immunity to infection, or which exacerbate disease through inflammation. Inflammation is a known factor in cancer, heart disease, arthritis, viral infection, and many other disorders. These systemic diseases in particular are difficult to study in vitro or in animal models and may benefit the most from the development of these chip-based technologies.
Making Tissue Chips
Replicating a tissue’s physiologic state requires having the right cells in the right proportions for the desired functionality, plus culture conditions that induce their assembly into functioning organ units, all while promoting the cells’ maturation into the appropriate phenotypes.
For example, the heart contains multiple cell types, including cardiomyocytes, fibroblasts, endothelial cells, and resident macrophages in specific ratios, packed densely and under strain. Since the heart’s mechanics affect its response to stimuli, cardiac tissue-chips should therefore incorporate this strain. Implanting electrodes that mimic the action of pacemaker cells allows “training” cardiac tissue chips to beat at the strength and frequency exhibited by human heart muscle.
Similarly, bone tissue chips incorporate not only relevant bone cells, but the rigid scaffolding and mechanical compression forces that bone tissue experiences during such ordinary activities as lifting and walking.
The final piece of the puzzle involves cellular maturation. Cells harvested from a patient’s differentiated progenitor cells are ‘young’ and lack the characteristics of mature cells that make them vulnerable to disease triggers. Since researchers don’t have decades to wait, our protocols include culture conditions for maturing cells rapidly and reliably.
As we learned from efforts to bioengineer organs for transplantation, supplying functioning vasculature to ex vivo tissues is problematic. We have not, for example, been able to engineer large channels and sprouting microvasculature on the same tissue chip. Instead, our devices incorporate a large built-in microfluidic ‘artery’, into which microvasculature forms in connection through angiogenesis.
It should be stressed that all cellular components of tissue chips, including the microvasculature, can arise from a single patient-derived progenitor cell – making it a patient-specific model for realising the goals of PM.
Why Are We Interested?
Adapting cell-based assays from suspension cells to 2D culture, 3D culture, and beyond to ‘patient in a dish’ systems is not an exercise undertaken lightly. Each step involves experimentation, optimisation, and most importantly validation – all of which divert resources from core operations. The effort pays off, however, on several levels.
Biological relevance, a feature mostly lacking in 2D adherent cell cultures, is the first benefit. We know that when compared with in vivo systems, cells cultured in three dimensions differentiate and produce metabolites at levels and in ways similar to their tissues of origin, and much more faithfully than 2D cultures. A 3D disease model based on spheroids or organoids should therefore more closely replicate critical interactions in drug screening or disease modeling. Similar advantages become available for cell growth, gene/protein expression, and drug resistance.
Moving up in complexity from simple 3D cultures to organ-chips and patient-chips provides the additional advantage of multiplexing, specifically the ability to obtain response readouts from either multiple cell types (organ-chips) or two or more organ systems (human-chips) at the same time, from the same stimulus, in the same experiment. In the past, obtaining this information required multiple animal studies or cell-based assays. Moreover, induced pluripotent stem cell technology allows generation of organ- and patient-chips comprised of multiple cell types with identical genetic signatures.
Combined, biological relevance and multiplexing provide cell-based assay platforms that may be engineered to quantify rather subtle responses, for example cell death in one cellular compartment and uptake of a drug in another. Moreover, once designed and validated, 3D cell culture systems and even patient-chips may be produced consistently and in large quantity, almost as reagents. This will facilitate collaborative science and perhaps mitigate to an extent the irreproducibility problem in the life sciences.
Our group has been actively working towards these goals as part of the NIH and NCATS Tissue Chip program. Guided by the collective effort to advance these technologies forward for more accurate screening of drug toxicity and disease efficacy, our group focuses on demonstrating the predictive ability of integrated tissue chip models in recapitulating patient specific responses. We bioengineered a millimeter-scale, multi-tissue (heart, liver, bone, skin, vasculature) platform in which individual tissues matured in custom-designed chambers and were connected through microfluidic vascular perfusion. Individual tissues are formed in a variety of ways, all designed to mimic their biological niche in vivo. For example, liver tissues are generated by combining hepatocytes and fibroblasts in 3D spheroids, which can be done using Corning Ultra-Low Attachment (ULA) microplates, and further combined into a biomimetic 3D hydrogel. Skin tissues are formed by encapsulating a series of cells into a hydrogel grown on a permeable insert, such as the Transwell membranes from Corning Life Sciences, so that they can be exposed to an air liquid interface in order to mature the skin into stratified layers. To further recapitulate the cells’ natural environment we introduced macrophages, also via perfusion. Using engineering approaches, we can connect multiple tissue chips in a way that preserves the long-term maintenance of biological integrity of these systems. All cells can be derived from a single, patient-derived induced pluripotent stem cell for patient specificity. In this model, tissues connected to one another show increased prediction of clinically observed drug responses and offer a systemic model of a patient’s disease. Overall, we are just beginning to realise the potential of these tissue-chip technologies and their utility for advancing PM.
Modern tissue chip models, which exploit microfluidics, 3D cell culture, and patient- or species specificity, will help accelerate drug development by resolving discrepancies in drug safety and efficacy observed among animal models, cell-based assays, and clinical studies. Tissue chips therefore serve, uniquely, as proxies for in vivo studies at both ends of the clinical development spectrum. At the earliest preclinical and Phase 1 stages, where one hopes eventual drug failures will occur, tissue chips provide both organ functionality and species specificity. During Phase 3 and post-marketing, they will assist in stratifying patients for trials or predicting both efficacy and toxicity in multiple organ systems.
Tissue chips cannot solve every problem related to human-like ex vivo models. Unlike spheroids, tissue chips focus on recapitulating select functional readouts per ‘organ’, so a multi-functional assay will require the design of additional chips for additional functions that may not be already represented. Also remember that complexity is a double-edged sword. The more functions built into the model the greater the effort in preparing it, the less likely it will prove to be robust, reliable, and producible in sufficient quantities, and ultimately the more involved the validation effort.
Volume 21, Issue 4 – Fall 2020
Dr Kacey Ronaldson-Bouchard is an Associate Research Scientist at Columbia University’s Department of Biomedical Engineering in New York City. Her research focuses on engineering multicellular systems from pluripotent stem cells, including the advanced maturation of iPSC-derived cardiomyocytes into mature, functional cardiac microtissues. Dr Ronaldson-Bouchard has an interest in integrating multiple engineered tissue models for studying inter-organ interactions, towards the development of patient avatar models for the realisation of personalised medicine. She has published papers in journals including Nature, Cell Stem Cell, andCell, and is a co-founder of TARA Biosystems, a NYC-based start-up company focused on engineering advanced cardiac models for drug discovery, testing, and precision cardiology.