Ex-Vivo Expansion of Stem Cells: Inspiration From a Highly Specialized Biologic Niche

1. Pediatric Hematology-Oncology Fellow, Oregon Health and Science University and Doernbecher Children’s Hospital

2. Associate Professor, Pediatrics, Oregon Health and Science University, Doernbecher Children’s Hospital

Background

An increasingly detailed understanding of hematopoietic stem cell (HSC) biology along with improved outcomes after transplantation of patients with malignant and nonmalignant diseases has fueled interest in HSC expansion. Clinically, both the limited availability of suitable HLA-matched donors and the fact that cord blood units contain a suboptimal number of HSCs remain obstacles that contribute to transplant-related morbidity and mortality as a consequence of graft-versus-host disease, delayed engraftment, and graft rejection.¹  Expanding allogeneic stem cell grafts in vitro prior to transplant is an attractive solution to overcome these barriers to successful transplantation. Moreover, such strategies may be useful for expanding autologous cells in the context of gene therapy for patients with heritable hematopoietic diseases. In the field of regenerative medicine, the capacity to expand stem cells ex vivo could lead to the generation of novel off-the-shelf products such as universal donor erythrocytes or genetically engineered HLA-compatible stem cells.^2,3  Techniques used for HSC expansion, once narrowly focused on transcription factors involved in HSC growth and renewal, have transitioned to a more integrative approach that incorporates extrinsic regulatory components that act in concert to regulate stem cell homeostasis. This integrative approach is driven by the results of insightful experiments that have identified and characterized the highly specialized elements of the bone marrow stem cell niche. Here we review key aspects of HSC biology and highlight promising approaches to HSC expansion.

Stem Cell Biology

The unique capacity for both self-renewal and multi-lineage differentiation is the mechanism by which HSCs maintain the supply of all mature blood cells throughout life (Figure 1). Ethical and practical concerns preclude extensive human research into the mechanisms that determine the direction of HSC division in vivo, but mouse models have been developed that offer insights into this remarkable process. In adults, under homeostatic conditions, HSCs exist predominantly in a dormant state with self-renewal being achieved through asymmetric outcomes from cell division, in which an HSC divides and produces one daughter HSC and one lineage-committed early progenitor cell (Figure 1A). Under specialized circumstances such as during physiologic development in utero, symmetric division dominates. In this case, two identical daughter HSCs arise from each division resulting in a 38-fold increase in HSCs by mid-gestation (Figure 1B*). Serial transplant studies further demonstrate the capacity of HSCs to expand through symmetric division. Such experiments have shown that a single HSC can expand 8,400-fold1 and multi-lineage reconstitution of hematopoiesis has been demonstrated in a xenograft model in which mice were infused with a single purified human HSC. In aggregate, these observations demonstrate the capacity of HSCs to expand and provide an experimental method by which to validate that HSCs that are expanded ex vivo retain the property of stemness by testing their capacity to repopulate the bone marrow in vivo.

Intrinsic Targets

Many early efforts targeting expansion of HSCs focused on manipulating genes involved in self-renewal. As an example, overexpression of the homeobox gene HoxB4 in murine HSCs via retroviral transduction yielded a 50-fold increase in HSC number over controls, boosting both short- and long-term repopulating capacity.⁴  Similarly, another transcriptionfactor, SALL4, has been shown to regulate pluripotency and, when made constitutively active in human CD34-enriched cells by way of lentiviral transduction, yielded a 118-fold increase in HSCs with long-term repopulating capacity in a xenograft model.⁵  However, while providing strong proof of concept for feasibility, approaches to HSC expansion that rely on commonly used retroviral vectors generally lead to sustained and unregulated overexpression of the transduced gene that, together with concerns about insertional mutagenesis, raised the daunting specter of treatment-related leukemogenesis.⁶  Accordingly, alternative approaches designed to extrinsically regulate both the level and duration of target gene activation so as to achieve more physiologic HSC expansion have become the focus of investigation.

Targets of Extrinsic Activation

Constitutive activation of the Notch signaling pathway in murine HSCs was shown to give rise to immortalized pluripotent cells capable of multi-lineage repopulation in vivo.⁷  As expression is physiologically regulated through the extracellular-binding domain of one of four receptors, investigators turned their attention to ligand-mediated, reversible activation of Notch. Induction of endogenous Notch expression in human HSCs using an immobilized form of the Delta 1 ligand yielded a 200-fold increase in the number of CD34-positive progenitor cells, and the expanded cells were shown to efficiently repopulate hematopoiesis in a xenogeneic transplantation assay. Results from the initial 10 patients enrolled in a phase I double cord unit transplantation trial showed a 164-fold Notch-mediated expansion of HSCs that resulted in a significant reduction in time to neutrophil engraftment.

Cellular copper has been implicated in HSC regulation and proliferation⁸  prompting investigation of copper as a target for extrinsic manipulation to expand HSCs ex vivo. Culturing human bone marrow cells with the copper chelator tetraethylenepentamine (TEPA) yielded a 17-fold and 159-fold increase in immunophenotypically defined human HSCs at week three and week seven, respectively, and the cultured cells were shown to engraft in murine xeno-transplantation assays. A phase I clinical trial utilizing TEPA-cultured cord blood cells demonstrated safety and feasibility and led to initiation of a phase II/III multicenter international trial, the results of which are not yet published.

Cytokines

Almost uniformly, the above studies utilized supportive cytokines during ex-vivo culture. For example, the effect of SALL4 was dependent on culture with stem cell factor (SCF), Flt-3 ligand, and thrombopoietin (TPO).⁵  Likewise, efficient Notch activation required the presence of various cytokines, including SCF, IL-3, IL-6, IL-11, Flt-3 ligand, and TPO.⁷  Not surprisingly, angiopoietins and cytokines, alone and in combination, have been investigated as potential mediators of HSC expansion. Human cord blood cells expanded with SCF, TPO, and granulocyte colony-stimulating factor (G-CSF) were found to be safe when infused in patients, but treatment yielded only modest expansion and provided no advantage in the time to neutrophil engraftment.¹  More recently, combinations of insulin-like growth factor 2 (IGF-2) or IGF-binding protein 2 (IGFBP-2) with angiopoietin-like proteins (Agptl), including Agptl2, Agptl3, Agptl5, and Agpt7, have been found to enhance ex-vivo expansion of stem cells up to 30-fold.¹  The observation that cytokines and growth factors that are produced by a variety of cell types greatly influence stem cell physiology suggests the importance of intercellular crosstalk and supports investigation of approaches to HSC expansion that mimic the HSC environment.

Stem Cell Niche

Elegant studies have explored the intricate relationship between stem cells and their specialized regulatory microenvironment in the bone marrow (and to a lesser extent the fetal liver).⁹  Long-lived HSCs appear to reside primarily at the endosteal surface of the niche where their properties are influenced by signaling from a variety of non-hematopoietic cells (Figure 2).⁹  Osteoblasts have been reported to play a critical role in HSC function through both paracrine regulation and direct cell contact. As an example, in mice, the number of osteoblasts expressing the cellular adhesion molecule N-cadherin correlates with the number of long-term repopulating HSCs, implicating N-cadherin in homing following bone marrow transplantation and suggesting that signaling via direct cell contact is a homeostatic process.¹⁰  Similarly, endothelial cells (ECs) appear to impact stem cell biology as experiments have shown that co-culture of human HSCs with human ECs results in a 400-fold expansion of immunophenotypic stem and progenitor cells.¹¹  Mesenchymal stem cells (MSCs) have also been found to enhance expansion of HSCs and inhibit differentiation during ex-vivo co-culture, findings that were dependent on cell-cell contact.¹²  Clinically approved protocols for extraction and propagation of MSCs exist, and investigators are now utilizing HSC–MSC co-cultures for clinical ex vivo expansion of cord blood-derived and peripheral blood stem cells. Reports suggest that this strategy can provide an approximate 40-fold expansion of immunophenotypic HSCs while decreasing the time to neutrophil engraftment. Interestingly, the role of the microenvironment is not limited to cellular interactions, as components of the intercellular matrix have also been implicated in the regulation of the stem cell niche. In particular, covalently bound fibronectin appeared to support the expansion of cord blood HSCs, an effect that was further enhanced in the setting of a 3D-synthetic scaffold.¹³  Similar findings with other extracellular matrix components have led to the development of synthetic biosubstrates modified with adhesion molecules in an attempt to mimic the structure of the stem cell niche.¹ 

Future Directions

Technologic advances in the development of high throughput screens of small molecule libraries have led to the discovery of several compounds with apparent HSC regulatory potential. For example, the HSC regulatory properties of prostaglandin E2 (PGE2) were first recognized as a result of high-throughput screening in a zebrafish model. Subsequent studies showed that PEGE2 expanded long-term repopulating HSCs in a murine transplantation model,¹⁴  and PGE2 is currently being investigated in clinical trials of ex vivo HSC expansion in humans. A different highthroughput screen identified an aryl hydrocarbon receptor antagonist (AhR) that yielded a 17-fold increase in cells with repopulating capacity in a mouse transplantation assays.¹⁵  This novel small molecule, now marketed as StemRegenin1 (SR1), is undergoing clinical trials in humans. In many cases, the mechanisms by which the identified candidate molecules modulate stem cell behavior are unknown; however, the discovery efficiency offered by high-throughput screening makes this technique an attractive tool for identifying molecules with stem cell regulatory potential that can be functionally validated subsequently using conventional culture assays. The future of stem cell expansion will likely incorporate biologic and bioengineering systems approaches that topologically mimic the stem cell microenvironment. Here microscopy and imaging are invaluable tools for guiding in vitro reproduction of the architecture of the stem cell niche. Clearly, understanding the molecular pathways that mediate HSC expansion through self-renewal without lineage commitment is central to harnessing the therapeutic potential of HSC expansion. Progress in the area of HSC expansion continues to accelerate, and success in the clinic will not only enhance the field of HSC transplantation, but will also fuel progress in gene therapy and regenerative medicine.