Advances in cell culture techniques don’t just mean more accurate models, but more personal ones. For decades scientists have been trying to figure out how to grow cells, especially diseased ones, directly from patients, not only to better understand the mechanisms at play, but also to determine the best approach to cure or correct the malfunction. Today, scientists can grow human tumors in a variety of designer mice, some with human immune systems. They can also grow cells directly from patients as either spheroids or organoids, pushing the boundaries of disease modeling and drug discovery. As such, treatment intersects with the individual; precision medicine isn’t on the horizon, it’s here.
In simplest terms, organoids are 3D structures that can be established from stem cells including pluripotent embryonic stem cells (ESCs) and newer generation induced pluripotent stem cells (iPSCs), and adult stem cells (ASCs). Ideally, the organoid or mini-organ possesses the full complement of cells that would be found in the body’s organ, and develops in such a way so as to emulate both architectural and functional characteristics of the desired organ. In theory, these structures should be amenable to standardization since the cell sources can be expanded, although there are stark differences between iPSCs and ASCs. Further, some tissues are relatively simple to reconstruct, such as the gut, whereas others will require more complex cocktails of growth factors and precise conditions to coax the construction of the model organ.
It all begins with collecting cells from a patient. “It could be a biopsy from the organ of interest such as a colon sample or pancreatic sample,” says Hilary Sherman, senior applications specialist at Corning Life Sciences. Elizabeth Abraham, senior product manager at Corning, notes that the personalized medicine model first relied on “adult tissue specific stem cells and organoids generated from diseased tissue. Now it is being extended to organoid models derived from iPSCs.” (Sherman says iPSCs are typically generated from somatic cells such as blood or skin.) Abraham continues, “The thinking is that these cells may also carry the mutational signatures. In the next decade, there will be more cross-talk between gene editing and personalized medicine, making more interconnected model systems, more targeted drug-screening efforts, and larger biobanks with a wealth of DNA-based data available to be mined.”
Extracellular environment critical to organoid growth
These cells require specific media formulations and an extracellular matrix (ECM). Corning supports “patient in a dish” by recreating the complex ECM microenvironments offering scientists Matrigel matrix for organoid cultures and Matrigel matrix-3D plates, says Sherman.
John O’Neill, chief scientific officer, Xylyx Bio, points out that this is a critical component of the patient in a dish paradigm. “A patient is more than simply the sum of their cells, and is defined in part by non-cellular components like ECM. Patient-specific approaches in research and diagnostic evaluations should consider not only cellular components such as patient-derived organoids but also environmental substrates that connect the patient milieu to the cells under evaluation. Xylyx Bio offers tissue and disease-specific extracellular matrix substrates representative of a specific tissue or disease profile.” These include ECM substrates derived from diseased tissue, for example, for liver and lung fibrosis. “Use of disease-specific ECMs can support a better understanding of patient-specific responses, and guide therapeutic development,” he adds.
Image: Xylyx Bio's library of tissue-specific ECMs
Tanya Yankelevich, director of product management, Xylyx Bio, notes that ECM substrates derived directly from tissue help not only to recreate the disease microenvironment for better modeling but allow for better assessment of compounds as potential tailored treatments. “Testing drugs in a specific disease environment will lead to improved outcomes through targeting treatment.”
Better disease modeling
Of all the diseases that require more personally tailored treatment options, Sherman says that cancer is a big one, since patients can respond very differently to chemotherapy. Tumor-derived organoids are similar to mouse patient derived xenograft (PDX) models in that they carry hallmarks of the original tumor and are predictive of a patient’s response to treatment. These are both powerful preclinical models. But organoids can be generated faster and are more easily scalable. As such, thousands of drugs can be screened to find the one most effective at killing the cancer. And patient-derived organoids (PDOs) can be especially useful for the more heterogeneous and harder to treat cancers that may be discovered at a later stage, including pancreatic and ovarian cancers.
For example, a recent study describes a platform for generating ovarian cancer PDOs that retain the features of the original tumors and can be used for drug sensitivity assays demonstrating the potential for personalized medicine. Further, clinical trials are now beginning to emerge using PDOs to establish clinical relevance, a first step in bringing personalized medicine to the clinic. Scientists at the Gustave Roussy Institute will examine their usefulness in guiding the treatment of patients with digestive cancer. This large, prospective study, Organotreat-01, is due to begin in 2021. Others will use PDOs to determine the best chemotherapy regimens for patients with lung cancer as well as pancreatic cancer.
Other diseases stand to benefit as well. Cystic fibrosis was the first disease to be successfully modeled in organoids in the laboratory of Hans Clever. Cystic fibrosis is caused by mutations in the CFTR (cystic fibrosis transmembrane conductance regulator) gene, which alters production, structure, or stability of the chloride channel. The efficacy of newer drugs that modulate this protein depend on the type of mutation carried by the patient. iPSC-derived mini-guts from biopsied cells can be used to predict which drug will be most effective. This is particularly useful for patients with very rare mutations. And Sherman points out that organoids can also be used for modeling infectious disease. Human and animal organoids are proving to be useful and effective modeling tools for both COVID-19 and Zika virus.
As protocols become standardized, and more labs confirm the use of PDOs as viable, sound modeling systems, matching the right drug to the right patient may no longer be a game of trial and error. Sherman thinks that more data will emerge demonstrating the value of organoid models. “I think the question will be if it is feasible on the scale necessary to implement.”