UC Berkeley

Stem cells are defined by two processes: self-renewal and differentiation. The balance of these two processes is determined by the signals--including extracellular matrix (ECM) proteins, soluble molecules, and other cells--in the microenvironment of the stem cell. To culture stem cells for cell replacement therapies, the microenvironment needs to support long-term self-renewal, including both the maintenance of multipotency or pluripotency and the proliferation, of the stem cells. Since defined medium conditions have been developed for many stem cells, including neural stem cells (NSCs) and human embryonic stem cells (hESCs), the focus of this work has been on the adherent substrates used during culture.

Many biomaterials--using natural, synthetic, or a combination of both types of materials--have been developed for neural stem cells. For example, natural materials such as collagen, other ECM proteins, and calcium alginate have been studied. Depending on the media conditions used, some of these natural materials promoted the self-renewal of neural stem cells. However, since most of these natural systems are not homogeneous because of the many isoforms of often impure ECM proteins present, synthetic materials have also been developed and investigated for NSCs. While synthetic materials have morphology and composition that is easier to control, in general synthetic materials lack the bioactive motifs necessary to actively engage and communicate with cells, resulting in low cell viability or premature differentiation. Accordingly, the field has been increasingly biofunctionalizing materials with motifs, such as peptides, that are capable of binding to cell adhesion recpetors, and studies with a biomaterial system utilizing an arginine-glycine-aspartic acid (RGD) peptide have shown that this material can support neural stem cells similar to their standard culture conditions.

Development of biomaterials for human embryonic stem cells (hESCs) has focused on finding natural and synthetic alternatives to Matrigel. Matrigel, a highly heterogeneous mixture of proteins including collagen IV and laminin, is still the typical adherent substrate for culture. Even though natural and synthetic materials have been explored as replacements for Matrigel, none of these materials has been shown to have the capacity for maintaining long-term self-renewal of hESCs similar to Matrigel. For example, materials containing RGD peptides do not support the growth of hESCs, as these cells utilize non-RGD-binding integrins for attachment to adherent substrates such as Matrigel-coated surfaces. Thus, there is a need to develop a method to find novel peptides that attach to a cell population such as human embryonic stem cells. Once developed, this method could find candidate peptides that could be used in biomaterials to replace Matrigel.

Since neural stem cell culture is supported by materials using a RGD peptide, this cell type was used as a model to develop and validate a method for finding novel peptides. The method developed for finding novel peptides included using selections with an unbiased bacterial peptide display library to find candidate peptides followed by further characterization of synthetic versions of some of the peptides for their ability to support the culture of neural stem cells. Using this general method with adult neural stem cells (NSCs), 44 high-binding bacterial clones were found. Of these clones, four contained RGD motifs commonly found in integrin binding domains, and three had homology to extracellular matrix proteins. Three synthetic analogs of peptides were chosen from the biomimetic ligand selections, grafted onto interpenetrating polymer network (IPN) surfaces, and adsorbed on tissue culture polystyrene (TCPS). These three peptides were found to support cell proliferation to different extents, but all three supported self-renewal of the NSCs on IPN surfaces, while all three peptides supported both proliferation and self-renewal when adsorbed on TCPS. This library-based approach, unbiased towards any particular motif, was shown to yield peptides that supported the culture of neural stem cells and that contained motifs that are known to bind to cell adhesion receptors, such as integrins. Now that our method using bacterial peptide display selections was developed, we applied this method to human embryonic stem cells.

Using the method developed for neural stem cells, many peptides were found that bound with high-affinity to hESCs. When four of these peptides were adsorbed on TCPS, one peptide supported the short-term self-renewal of hESCs as indicated by proliferation and maintenance of pluripotent markers. In addition to finding general cell-binding ligands, selection was then targeted to a particular adhesion receptor, in this case the α6β1 integrin, to recapitulate engagement with laminin. From the targeted selections, many peptides with high affinity for hESCs were found. Of the five tested when adsorbed on TCPS, two of them supported short-term self-renewal of hESCs. Overall, the development of a method utilizing bacterial peptide display selections to find novel peptides successfully found peptides that supported the culture of neural stem cells and human embryonic stem cells, with the best performing peptides being obtained from selections targeted for the α6β1 integrin on hESCs.