The goal of this research is to create in vitro microenvironments for long term culture of hematopoeitic stem cell (HSC) in microfluidic bioreactors. In vivo, HSCs reside in the bone marrow osteoblastic and vascular niches in adult mammals. Defining features of their in vivo niche include: small number of HSCs, heterogeneous population of bone marrow cells that support HSCs, and low oxygen tension. We engineer niche elements in microfluidic bioreactors by modulating oxygen tension, optimizing attachment and growth of HSC-supporting bone marrow stromal cells, and culturing small numbers of HSCs in their physiologically relevant ratios between HSCs and supporting cells.

By using a combination of a mathematical model and quantitative experiments, we have created a design tool to manipulate and control oxygen tension for cell culture inside the poly(dimethyl siloxane) (PDMS) microbioreactors. Dissolved oxygen concentrations in the microbioreactor are quantified in real time using fluorescence lifetime imaging of an oxygen sensitive dye. Experimental results are consistent with the mathematical model and give insight into operating conditions required for a desired oxygen tension in cell culture regions of the microbioreactor.

We used microfluidic perfusion systems to develop nanocoatings made from electrostatic self assembly of PDDA (poly(diallyldimethyl ammonium chloride)), clay, type IV collagen and fibronectin to optimize attachment of primary murine bone marrow cells (support cells for HSCs) onto PDMS bioreactors. PDDA-topped coatings were found to be cytotoxic, while coatings with two or more bilayers of proteins collagen and fibronectin were found to optimize spreading, proliferation, and viability as compared to other surfaces.

On-chip erythropoiesis was achieved with a 3-D co-culture of HSCs with supporting cells in PDMS bioreactors. In addition, an optimal ratio of support cells to HSCs was found to maximize self renewal potential of HSCs in vitro. By the combination of hypoxia (which simulates in vivo bone marrow oxygen tension), biofunctional surfaces, and 3-D co-cultures, we are moving towards a 'microfluidic HSC niche', in which hypothesis-driven studies about crosstalk between HSCs and stromal cells can be carried out.