Manipulation of the extracellular microenvironment by micro- and nanotechnology approaches to improve the generation of pancreatic endocrine cells from human embryonic stem cells

Abstract

Human embryonic stem (hES) cells have the ability to generate all cell types in the body, which suggest that they can provide an unlimited source of cells for cell replacement therapy to treat degenerative diseases such as diabetes mellitus. To achieve a stem cell therapy treatment for diabetes mellitus, the hES cells must be differentiated into mature functional insulin producing beta-cells. Current differentiation protocols focus on the addition of soluble molecules whereas the impact of the physical microenvironment has been mainly unattended. However, the physical microenvironment plays an essential role in cellular behaviour during development and recent studies have demonstrated the effect of the physical environment in in vitro stem cell differentiation. Thus, understanding the role of the physical microenvironment is vital for the development of effective and consistent differentiation protocols. In this study we manipulated the physical environment during the first two differentiation steps towards beta-cells; definitive endoderm (DE) and pancreatic endoderm (PE). Three different approaches were used: 1) systematically screening of extracellular matrix (ECM) substrates (paper 1), 2) alteration of substrate topography and elasticity (paper 2) and 3) initial seeding and cell distribution (paper 3). With the first strategy an array screen was performed to systematically identify ECM protein coatings, which induced or inhibited the differentiation of hES cells towards DE. Almost 500 different ECM protein combinations were screened and several candidates were found. The majority of these candidates could be validated in microtitre well plates and further studies demonstrated that certain ECM proteins regulate the differentiation of hES cells towards DE. Netrin 1, collagen 1 and collagen 2 induced DE differentiation to a higher degree than the control fibronectin. Especially, all the analyses pointed to collagen 1 having unique properties. Cultures on collagen 1 had distinct morphology, proliferated faster and most importantly resulted in purer DE cultures with very few undifferentiated cells. Currently, the underlying biological mechanism is not known and is a subject for further studies. However, to our knowledge collagen 1, collagen 2 and netrin 1 have previous not been linked to embryonic stem cell differentiation. Notable, this study demonstrated that the ECM proteins do have a functional role in stem cell differentiation and should be taken into consideration in order to obtain efficient and consistent differentiation protocols. With the second strategy the physical microenvironment was manipulated with topographies in micro and nanoscale which have different elastics characters. The study consisted of an investigation of the differentiation of hES cells towards DE and PE on nanopillars. The nanopillars with soft character were not favoured by hES cells with regards to attachment and growth. However, DE cells appeared earlier during the DE differentiation on the soft nanopillars when compared to the flat control surface. Moreover, DE cells intensively repopulated the area with soft nanopillars, whereas the undifferentiated cells remained on the flat surface area. The very fast repopulation of DE cells indicated DE cells migrated onto the soft nanopillars, implying that the DE differentiation could be durotaxin driven, where the differentiated cells migrated from rigid substrates towards soft substrates. In contrast, the differentiation towards PE was to a large extent repressed on the soft nanopillars in comparison to the flat control surface. This indicated that the requirements of the physical environment changes during the different stages of differentiation. Our study demonstrated that the differentiation and cellular behaviour of embryonic stem cells are affected by altering the physical properties of the cell culture surface with nanopillars. Such observations have previously not been reported with embryonic stem cells, indicating that additional dimensions, such as the physical environment, should be taken into account when directing stem cell differentiation. With the last strategy, the cell seeding density and cell distribution across a well was investigated. With a simple cell seeder device, an even and consistent distribution of cells across individual experiments was obtained for human fibroblast cells, hES cells and DE cells derived from hES cells. A uniform seeding of DE cells resulted in a more uniform differentiation towards PE across the entire well (12-well plate). Such uniformity is important in reproducibility and for assays which includes material from the entire well.