Objective
The objective of the proposed work is to understand and mitigate the inherent, biological sources of variability associated with using human cells as a raw material for manufacturing, to decrease or eliminate that variability by more precisely guiding stem cells through their conversion to SC-islets, and then to demonstrate the consistent, automated, scaled-up production of SC-islets from stem cells from three patients that have banked stem cells at the Joslin Diabetes Center in the hopes that those cells might be used to cure their disease. Specifically, we will gain understanding the variability associated with cells through the use of AI, which will help us to sort through the thousands of biomolecules contained within and secreted by human cells to know which ones we need to measure during the SC-islet production process so that functional SC-islets are generated. We will then perform scaled-up production of islets we have characterized well to understand what differences arise during stem cell conversion to islets or in the islets themselves as a result of the use of manufacturing equipment and cell culturing equipment that is different from what is typically used in the research laboratory. Finally, we will apply the in-depth understanding of the SC-islet production process to the scaled-up automation of cells banked by the Joslin patients.
Background Rationale
The Millman laboratory has had success in generating islets from stem cells that are highly efficacious in reversing laboratory-induced diabetes in mice. Although these islets are generated at scales suitable for the reversal of diabetes in mice, we will require many orders of magnitude more islets to treat the significant numbers of Americans suffering from T1D. Furthermore, similar to other laboratories studying stem cell-derived islets, Dr. Millman’s laboratory has been successful at the highly efficient conversion of stem cells to islets from only a few stem cell sources, not always at high yield, and the biological rules governing the conversion of stem cells to SC-islets is not well understood. As a first aim of the proposed research, we will measure and empirically derive the rules governing the conversion of stem cells to SC-islets, and then use that information to culture stem cells from a variety of sources so that they all generate SC-islets successfully, as measured by their ability to secrete insulin in response to being exposed to glucose, similar to how they function in the body in healthy individuals. In order to scale up production to clinically-relevant levels, Dr. Millman’s laboratory recently collaborated with the Advanced Regenerative Manufacturing Institute (ARMI), who have expertise in the development of scalable, automated cell and tissue assembly lines, called “Tissue Foundries”, to scale up production of islets from one of his stem cell lines. During the collaboration, ARMI successfully scale up production of SC-islets up to 100-fold from the scale performed in Dr. Millman’s laboratory and developed an automated assembly line to carry out the process. However, automation can only address variability in a process that arises from a person performing its operations. Because stem cells are an inherently variable starting material, as evidenced by the difficulty with which stem cells from different donors are differentiated to SC-islets, we plan to further mitigate the challenges associated with variability by generating a deep understanding of its sources and then addressing them through biochemical means, for example through the addition of biomolecules to culture broths that allow the cells to grow. Once this variability can be addressed, then stem cells from a variety of sources will be amenable to scaled-up production on the Tissue Foundry.
Description of Project
Pancreatic islets from deceased donors have proven to be efficacious as a replacement therapy for Type 1 Diabetes (T1D) in numerous patients. However, the availability of these islets is limited and a new source of implantable islets is needed. The combination of decades of research and development efforts across several life sciences fields has led to our ability to generate various cell types, including those that constitute pancreatic islets, from human stem cells, which can be amplified significantly in number in culture and are therefore abundant. The significance of the potential application of stem cell-derived islets (SC-islets) as a replacement therapy for diabetes was most recently demonstrated by the successful treatment with SC-islets of a patient by Vertex Pharmaceuticals. However, the difference between generating sufficient numbers of SC-islets for a single patient in a laboratory and generating sufficient numbers of SC-islets to treat the more than 1.9 million Americans currently suffering from T1D is equivalent to the difference between cooking a pot of chicken noodle soup at home on a stove and the automated consistent production of millions of cans of chicken noodle soup in a Campbell’s factory. Even more daunting, the large-scale production of SC-islets requires us to understand the fundamental characteristics of human cells, which are subject to significant variability based on the identity of the donor, how they were originally isolated from human tissue, and the type of equipment being used to culture them. Therefore, we propose to understand this variability by making extensive measurements of the thousands of potential characteristics typical of living cells along the SC-islet manufacturing pathway and to use those measurements, in conjunction with the modern and emerging tools of artificial intelligence, to understand which subset of those characteristics need to be monitored closely so that production always leads to a batch of SC-islets that will be safe and functional in a patient, regardless of the source of cells or how they are manufactured. We will then monitor that subset of characteristics during laboratory-scale and scaled-up, automated production of SC-islets on a manufacturing line we have developed, and will ensure they are exhibited by the cells during manufacturing through the timely addition of specific biomolecules to the culture broth in which the cells are grown. Finally, we will demonstrate that the application of these tools leads to the consistent SC-islets by automating scaled-up production from stem cells derived from three individual patients by the Joslin Diabetes Center.
Anticipated Outcome
In the short-term, we anticipate we will generate sufficient numbers of SC-islets to support ongoing research and development activities into islet development, the etiology of diabetes and the study of its treatment, and for the development of technologies that minimize or eliminate the need for immunosuppression. In the long term, we anticipate that we will develop procedures and manufacturing infrastructure to facilitate the consistent, cost-effective production of stem cell-derived pancreatic islets at scales appropriate for the treatment of the T1D patient population in the United States and then in the world. These procedures will impact both the large scale production is SC-islets from a single donor and for personalized production, where individual patients are the donors of stem cells for the production of their own islets.
Relevance to T1D
Transplantation of insulin-producing cells and tissues promises to be a ‘functional cure’ for Type 1 Diabetes but is limited due to the scarcity and variation of available cell sources from deceased donors. Success of this research will lead to new cell products that could be transplanted into patients with type 1 diabetes, with these exogenous cells replacing the function lost in the endogenous beta cells, and control the blood sugar levels of patients without insulin injections or constant blood sugar monitoring. The goal is to replace absent or dysfunctional endogenous beta cells with beta cells derived from hPSCs that are functionally sufficiently similar to endogenous beta cells to normalize glycemic and metabolic control in individuals with T1D.