Exploration of a 3D Scaffold Culture System for Ex Vivo Blood Production

The human body produces 2.5 million red blood cells (RBC) every second in the bone marrow, where differentiation of immature erythroblasts to reticulocytes occurs largely within erythroblastic islands. Erythropoiesis can be routinely replicated in 2D liquid culture using hematopoietic stem cells (HSC) isolated from adult peripheral blood. However current 2D culture methodologies eventually exhaust the input of HSC. The use of 3D scaffolds to better mimic erythropoiesis in the bone marrow would increase the RBC yield and longevity of ex vivo cultures, whilst reducing use of exogenous cytokines and minimizing handling requirements.

As a starting point we are utilizing a porous collagen coated synthetic polyurethane (PU) scaffold (0.5cm by 0.5cm) for static 3D cultures provided by the Mantalaris group at Imperial College London (Mortera-Blanco et al., 2011). The PU scaffolds were seeded with lineage depleted or CD34⁺ population isolated from adult peripheral blood and maintained in serum-free erythroid expansion media with SCF, IL-3 and Dexamethasone, alone, or with erythropoietin (EPO) or thrombopoietin (TPO). Advantages of using the lineage depleted population is that it provides a larger diversity of stem cells for establishment of the niche, potentially facilitating the use of rare patient blood samples which may only be available in small sample volumes with low numbers of CD34+ cells. Scaffolds are productive using both cellular inputs, with significant cellular egress for up to 5 weeks regardless of whether exogenous EPO or TPO were included. As anticipated, the highest increase in cell production from the scaffolds was observed using CD34+ in the presence of EPO, which also provided a significant reduction in cell death. Histology and immunofluorescence were used to explore the cell populations within the scaffold. No mature macrophages were detected but GPA+ cells within the scaffold was observed until the end of culture, suggesting that cellular expansion is occurring without establishment of the classical macrophage niche.

We also characterized the cells that continually egress from the CD34+ scaffold cultures using flow cytometry. Typically the cellular output exhibited approximately 20-60% CD34 positivity dropping to <15% post day 24 of culture and 10-30% GPA positivity, the remainder of the population was largely CD61 positive, with a contingent of CD14 positive monocytes. Mature cell surface markers for erythroid and megakaryocytic lineages were detected in up to half of the population, when the harvested 3D egress cells were further cultured in erythroid or megakaryocyte 2D culture systems. This suggests a high proportion of the cells that egress from the scaffold are megakaryocyte erythroid progenitors (MEPs) that are consistently expanding within the scaffold environment for the entire culture period.

Since CD34+ expansion here has been achieved in the absence of macrophages, we wanted to explore their effects in 2D with the intention of introducing macrophages or macrophage derived functionality into our next generation scaffolds. Importantly, macrophage inclusion significantly increased proliferation of expanding erythroblasts compared to erythroblasts alone. Expansion of CD34⁺ cells in co-culture gave a statistically significant average fold increase of 528 compared to 301 for the control at day 7 (p = 0.0126 and 0.0162 for days 5 and 7 respectively (n=5)). Flow cytometry at the endpoint of the experiment showed a larger CD34⁺ population and a reduced GPA⁺ population when cells are in co-culture, suggesting the CD34⁺ cells are maintained in a more primitive state for longer. Therefore, co-culture has the additional benefit of improving early erythroblast expansion, alongside the recently reported enhanced expansion of erythroblasts during terminal differentiation (Ramos et al., 2013).

In summary, we have demonstrated that a basic static PU scaffold can be utilized to increase hematopoietic stem cell culture longevity and facilitate generation of megakaryocyte or erythroid progenitors with expansion potential. This occurs in the absence of any detectable macrophage niche generation. We have also shown that co-culture with macrophages enhances erythroblast expansion in 2D. Further work is needed to determine whether inclusion of macrophages or macrophage derived proteins in our scaffolds will effectively boost progenitor production.