Exploring the ability of mesenchymal stem cells to sense mechanics of 3D environments
04 October 2016
Tissue engineering is an exciting area of research that aims to use cells, biomaterials and biochemical factors to replace or repair damaged tissue. Advances in tissue engineering may provide a means of treating damaged tissues such as cartilage for which currently the only viable treatment is replacement with artificial prosthesis.
Additionally, continued progression of tissue-engineering technologies may eventually overcome the shortage of donor tissue available for organ transplants. A key aspect of future advances is the ability to control the activity of the cells that are used to build new tissue. Mesenchymal stem cells (MSCs) isolated from the bone marrow of patients or donors hold great potential as a cell source for tissue engineering. This arises from their ability to differentiate (or transform) into any one of several different cell types.
For example, MSCs can become bone-forming cells and subsequently be used to produce bone grafts. Naturally, the ability to control the differentiation of these cells is vital for successful clinical translation.
Traditionally, biochemical factors are added to the liquid media that cells are sustained with in order to control their differentiation. However, more recently, there has been much interest in the ability of the mechanical properties of the material that the MSCs are attached to in directing their differentiation. Indeed, it has long been known that mechanical loading is important in directing cell behaviour and tissue formation.
The increase in bone mass in the racket arm of tennis players is a good example. With MSCs, we find that cells attached to soft materials tend to become fat-forming cells, while cells on stiff materials tend to become bone-forming cells. This ability of cells to gauge and respond to the mechanics of their surrounding substrate is termed mechanotransduction and much of our understanding of this area has been gained from experiments carried out on flat, 2D surfaces.
However, for tissue-engineering applications, a biomaterial scaffold is typically used to provide a 3D support or guide for the cells while they form new tissue. To this end, we pursued a collaborative project to develop materials and models that would allow us to probe the changes in cellular mechanotransduction as we moved from 2D to 3D.
This project started when Dr Matthew Haugh moved to Prof Sarah Heilshorn’s laboratory at the School of Engineering in Stanford, California, to learn about recombinant biomaterials as part of an Irish Research Council-funded ELEVATE postdoctoral fellowship. Prof Heilshorn is associate professor of materials science and, by courtesy, of chemical engineering and of bioengineering.
Recombinant proteins are a new class of biomaterial, which leverage advances in genetic engineering of bacteria to produce customised proteins (Fig 1). In essence, we are using bacteria to produce tailored versions of the building blocks that our own cells use to create all the tissues within our body. This allows us to make natural environments for the cells and also provide robust control over material properties (cellular attachment site chemistry and mechanics).
While at Prof Heilshorn’s laboratory, Dr Haugh developed methods to form porous scaffolds from these recombinant proteins using a sacrificial template of acrylic microspheres.
This technique enables us to explore the response of MSCs to various material cues in 3D environments, which is work we are currently exploring in Prof Fergal O’Brien’s group at the Royal College of Surgeons in Ireland (RCSI). Prof O’Brien is professor of bioengineering & regenerative medicine, deputy director of research and head of the RCSI Tissue Engineering Research Group.
During some of the first experiments exposing mesenchymal stem cells to these materials at Stanford, it was noticed that the cells were taking up unusual shapes within the hydrogels. Following these experiments, we began to think about how cell shape and the architecture of the 3D porous structures would affect the mechanical feedback that the cells sensed from the material (Fig 2A).
Upon discussion of this idea with Dr Ted Vaughan, we realised that computational modelling would be the ideal tool to explore this idea.
Cells probe the mechanical properties of their surrounding environment by generating contractile forces in their cytoskeleton, which essentially means that they have the ability to pull on the substrate to which they are attached. The deformation of the substrate in response to an applied force provides the cell direct feedback on the local stiffness at this location.
In the case of flat 2D surfaces, the mechanical feedback that the cell experiences depends only on the elastic modulus (E) of the substrate material. However, in the case of 3D environments, the mechanical interaction between the cell and the substrate becomes complicated by structural effects – in particular, the architecture of the scaffold in vicinity of attachment site (see Fig 2A).
This implies that the mechanical feedback experienced by cells changes from 2D to 3D environments, despite the fact the material’s elastic modulus may be identical. Due to the length-scale involved (~ 1μm), experimental measurements of local mechanical behaviour is extremely challenging and, therefore, Dr Vaughan sought to develop computational models to characterise local cell-scaffold interactions through finite element analysis (FEA).
Finite element analysis
These FEA models use an iterative approach to determine local cell-scale stiffness values across a range of 3D scaffold architectures, such as that shown in Figure 2B. The approach applies a series of unit displacements to microscale attachment sites located throughout the interior scaffold surface, with local stiffness values being determined from reaction forces generated during the simulation.
The resulting stiffness map of the scaffold surface (see Fig 2B) highlights the precise variation in local stiffness throughout the scaffold architecture. These local differences in mechanical response may have important implications for stem cell response, potentially affecting their migration, proliferation and, most importantly, their differentiation to more mature cell-types (e.g. fat or bone cells).
Interestingly, the results of this study also showed that scaffolds with relatively high porosities only saw modest reductions in the local stiffness sensed by cells. For example, for the 90% porous scaffold shown in Figure 2B, the measured local stiffness was approximately one-third of the material-level stiffness.
This reduction is relatively small, considering that the overall mechanical behaviour of this scaffold would be vastly compromised, with the bulk modulus seeing an approximate 20-fold reduction compared to a non-porous solid.
Overall, the modelling framework provides insight into the mechanical niche in three-dimensions and highlights that cells probe mechanical stiffness at a completely different length scale to traditional engineering measurements of elasticity
This work is a textbook example of how computational modelling and bench top research can complement each other to allow novel insights that would not be possible in isolation. Ongoing work in this area includes the application of the computational-modelling techniques to different scaffold geometries and experimental investigation of the relationship between biomaterial stiffness and attachment chemistry in guiding mesenchymal stem-cell behaviour.
Ted Vaughan, lecturer in biomedical engineering, NUI Galway
Matthew Haugh, IRC ELEVATE postdoctoral fellow at the Royal College of Surgeons in Ireland
Romano NH, Sengupta D, Chung C, Heilshorn SC. ‘Protein-engineered biomaterials: nanoscale mimics of the extracellular matrix.’ Biochim Biophys Acta. 2011; 1810(3):339-49