WO2007027226A2

WO2007027226A2 - Systems and methods for the production of differentiated cells

Info

Publication number: WO2007027226A2
Application number: PCT/US2006/016228
Authority: WO
Inventors: Krishnendu Roy; Sabia Taqvi
Original assignee: Board Of Regents, The University Of Texas System
Priority date: 2005-04-28
Filing date: 2006-04-28
Publication date: 2007-03-08
Also published as: WO2007027226A3

Abstract

The present invention generally relates to methods for the ex vivo expansion of undifferentiated cells. More specifically, the present invention relates to systems and methods of differentiating undifferentiated cells into a desired linage by providing to an undifferentiated cell stromal cell conditioned medium and a differentiation inducing ligand. The system of the present invention may be used to expand any undifferentiated cell that requires cell-derived soluble factors and cell-contact dependent signals. The present invention does not rely on transfected stromal cells as the signaling entity for the creation of differentiated cells.

Description

SYSTEMS AND METHODS FOR THE PRODUCTION OF DIFFERENTIATED CELLS

PRIORITY CLAIM

This application claims priority to United States Provisional Application No. 60/675,803 filed April 28, 2005, the content of which is incorporated in its entirety herein.

BACKGROUND OF THE INVENTION The process of cell differentiation is tightly regulated by soluble factors and cell contact-dependent signals within specialized microenvironments, each of which support the development of specific cell lineages. Accordingly, differentiating undifferentiated cells, like stem cells, into differentiated cells is often difficult. For example, differentiating stem cells into T cells often requires recombinant methods or methods that involve co-culture with thymus. Hematopoiesis is a good example of the differentiation process.

During hematopoiesis, cell mediated and soluble growth factors bind to cell surface proteins to trigger signal cascades, committing stem cells to either a myeloid or lymphoid lineage. Hematopoiesis usually occurs in the adult bone marrow microenvironrnent where supporting stromal cells encourage the growth and differentiation of progenitors through secretion of soluble factors and through cell-to- cell interactions. The meshwork of stroma and extracellular matrix (ECM) that make up the bone marrow provide not only the physical framework for hematopoietic cell proliferation and differentiation, but are also intimately involved in adhesion, cytokine presentation, and cell growth. The development of the various hematopoietic cell lineages is compartmentalized during fetal development and throughout adult life. At approximately day 12 of embryonic development, the fetal liver (FL) is seeded by definitive hematopoietic stem cells (HSCs), which arise from the aorta-gonad- mesonephros region of the developing embryo. The FL continues as the primary site of hematopoietic development until birth, when the bone marrow (BM) takes over as the primary site for hematopoiesis in the adult. Thymocyte development takes place in a complex milieu of supportive cells and ECM that are responsible for the proliferation, adhesion, migration, and selection processes these cells undergo before reaching maturity. T cell development is comprised of a series of complex interactions that take place both in the bone marrow and thymus and ultimately result in the formation of CD4+ and CD 8+ T cells that are capable of recognizing major histocompatibility (MHC) class II or class I molecules respectively, informing the immune system of extracellular or intracellular infections. Hematopoietic progenitors migrate to the thymus via the blood and undergo differentiation into T cells through specific and complex microenvironmental signaling. The expression of ligands and molecules in the thymic microenvironment are directly responsible for the proliferation, adhesion, migration, and selection of these progenitors during T cell maturation (Anderson et ah, Semin. Immunol., 2000, 12(5):457-64; Germain, Nat. Rev. Immunol, 2002, 2(5):309-22; Goldsby, Immunology, 2003, New York: W.H. Freeman & Co.). Several of the signals necessary for thymocyte development and survival including Notch signaling and MHC-TCR interactions have been characterized through over expression, gain-of- function, loss-of-function, and transfection studies (Germain, 2002, supra; Maillard et al, Annu. Rev. Immunol, 2005, 23:945-74).

Efforts to replicate one or more of these events in vitro have had to depend on the use of fetal thymic organ culture. The molecular interactions responsible for this thymus dependency remain largely unknown. However, a number of recent studies have implicated Notch receptor-ligand interactions in the earliest T cell lineage commitment events. Notch signaling is an evolutionarily conserved pathway that controls multiple cell fate decisions throughout ontogeny. Notch signaling is known to be well conserved throughout evolutionary development in a variety of organisms. The role of Notch signaling in the regulation of differentiation and self-renewal in systems such as hematopoiesis and myogenesis has also been well characterized (Maillard et al, 2005, supra; Varnum-Finney et al, Blood, 2003, 101(5):1784-9; Tan-Pertel et al, J. Immunol, 2000, 165(8):4428-36; Han et al, Blood, 200O₅ 95(5):1616-25; Karanu et al, J. Exp. Med., 2000, 192(9): 1365-72; Radtke et al, Nat. Immunol, 2004, 5(3):247-53; Parreira et al, Semin. Immunol, 2003, 15(2):81-9). Specifically in T cell lymphopoiesis, Notch signaling has been shown to be a necessary criterion. In the absence of Notch signaling lymphopoiesis only occurs along the B cell lineage, whereas the presence of Notch ligands on the surface of stromal cells provide signals necessary for T cell generation (Radtke et al, 2004, supra; Parreira et al, 2003, supra). Notch receptors and the Delta and Jagged families of Notch ligands are tightly regulated in their expression both in the bone marrow and thymus to achieve a unique balance of lymphocyte development (Maillard et al, 2005, supra; Radtke et al, 2004, supra; Parreira et al, 2003, supra). The expression of all four receptors by the developing thymocytes and the supportive thymic stroma has been previously characterized while expression of Notch ligands, Delta like ligand 1 and 4 (DLLl and DLL4), has been shown to occur in the thymic stroma (Parreira et al., 2003, supra). However, studies assessing the importance of cell-ligand ratio as well as duration of ligand presentation have not been explored in detail.

Current systems to generate T cells in vifro have largely relied on co- culture of stem cells with either fetal thymus isolated from mouse (fetal thymic organ culture, FTOC) or more recently with bone marrow derived stromal cells (OP9 cells) retro virally transfected with the Delta and Jagged families of Notch ligands (Hozumi et al, Nat. Immunol, 2004, 5(6):638-44; Hozumi et al, J. Immunol., 2003, 170(10):4973-9; Schmitt et al, Nat. Immunol, 2004, 5(4):410-7; Schmitt and Zuniga-Pfiucker, Immunity, 2002, 17(6):749-56; La Motte-Mohs et al, Blood, 2005, 105(4): 1431-9; Lehar and Bevan, Immunity, 2002, 17(6):689-92; Dallas et al, J. Exp. Med., 2005, 201(9):1361-6)> These systems, although effective, are complex, require genetic manipulation of stromal cells, and most importantly do not allow quantitative analysis of the effects of ligand to stem cell ratio or the duration of Notch signaling on T cell differentiation. Tissue engineering of T cells from progenitor populations in a high-throughput and efficient manner could benefit from synthetic approaches involving biomaterial-directed Notch ligand presentation.

Thus, a synthetic Notch signaling system using ligand-functionalized magnetic microbeads (artificial stromal cells) was used to evaluate how Notch ligands, specifically DLL4, presented through a biomaterial surface affect cell differentiation, and to develop a high throughput strategy to engineer lineage committed cells from undifferentiated cells; as exemplified, T cells from hematopoietic progenitor populations. The results show that stem cell-stromal cell contact is not a necessary event for undifferentiated cell commitment with differentiation inducing ligand. This system permits easy, high-throughput, scalable production of differentiated cells or undifferentiated cells poised for commitment because it avoids the need for intimate contact with stromal cells. Instead, it relies on contact with essential stromal cell paracrine factors and the differentiation inducing ligand on an artificial support, e.g., artificial stromal cells. To obtain pure differentiated cells, one need only remove the support. In this case, magnetic beads are a convenient support. Alternatively, pure undifferentiated cells seeded on a scaffold decorated with the differentiation ligand are suitable for implantation, where paracrine factors from tissue resident stromal cells assist with differentiation in situ. Such a bead-based artificial signaling system also allows quantitative study of the effects of ligand density and signaling duration, thereby providing further insights into the individual roles of the Delta and Jagged families of Notch ligands in T cell differentiation.

Accordingly, current techniques for in vitro T cell formation rely on cells transduced with a Notch ligand. Although Notch-ligand mediated signals have been shown to be a necessary component of T cell generation from stem cells, high- throughput, synthetic biomaterial-based systems for Notch-directed stem cell differentiation into lymphocytes are yet to be reported. Thus, there remains a need in the art to develop functionalized biomaterials and synthetic microenvironments to efficiently generate T cells from hematopoietic stem cells (HSCs). Eventually this could lead to the production of CD4+ and CD8+ cells that could be used for therapeutic purposes. In addition the current methods are not generally high throughput methods and are difficult to implement on a large scale. Thus, there is a need for a scalable approach to inducing differentiation of undifferentiated cells.

SUMMARY OF THE INVENTION The present invention generally relates to systems and methods for the ex vivo expansion of undifferentiated cells. More specifically, the present invention relates to systems and methods of inducing differentiation of undifferentiated cells into a desired linage.

The present invention provides for a system for inducing differentiation of undifferentiated cells, comprising a stromal cell paracrine factor, e.g., in a stromal cell conditioned medium; undifferentiated cells; and a differentiation inducing ligand attached to a support. In one embodiment, the system contains stromal cells, which can be a stromal cell line or primary stromal cells. The selection of stromal cell type depends on the differentiation commitment desired. For example, if one desires differentiation of neurons, a stromal cell able to support neuronal differentiation would be selected. In a specific embodiment, the stromal cells are bone marrow derived stromal cells, where the differentiation is into lymphocytes.

In a further embodiment of the invention, the stromal cells of the stromal cell conditioned medium are physically separate from the undifferentiated cells. In particular embodiments, this can be achieved through a physical barrier, or by separating the stromal cell conditioned medium from the stromal cells. Alternatively, the stromal cells can be in contact with the undifferentiated cells, although this complicates obtaining the differentiated cells in a more pure form. In a specific embodiment, the undifferentiated cells of the conditioned medium system are hematopoietic progenitor cells.

In particular embodiments of the system, the differentiation inducing ligand support is a bead or a scaffold. One specific embodiment provides that the bead is a magnetic bead. In an alternative embodiment, the scaffold is a bioerodable matrix, such as a PLGA matrix. A further embodiment provides that the differentiation inducing ligand is attached to the support via a biotin-streptavidin link.

A specific embodiment of the system provides that the differentiation inducing ligand and the stromal cell conditioned medium induce differentiation of a hematopoietic progenitor cell. A further embodiment provides that the hematopoietic progenitor cell differentiates into a T cell. In a further embodiment, the differentiation inducing ligand is a Notch ligand. In a further embodiment, the composition comprises a stromal cell paracrine factor, e.g., in a stromal cell conditioned medium.

The present invention provides for a composition of matter comprising undifferentiated cells; and a differentiation inducing ligand attached to a support. The undifferentiated cells, support, ligand, and stromal cell conditioned medium are as set forth above with respect to the system of the invention. One embodiment of the present composition of matter provides that the undifferentiated cells are hematopoietic progenitor cells. A further embodiment provides that the composition is an in vitro composition.

The present invention further provides a method comprising contacting an undifferentiated cell with (a) a stromal cell paracrine factor and (b) a differentiation inducing ligand attached to a support. The undifferentiated cells, ligand and support are as set forth above in connection with the system. The stromal cell paracrine factor can be present in a stromal cell conditioned medium, as set forth with respect to the system. Alternatively, a stromal cell paracrine factor can be supplied by cells in situ upon introduction of the undifferentiated cell and the differentiation inducing ligand attached to a support. A further embodiment of the method provides that the contacting is done in vitro. A further embodiment provides that the undifferentiated cells are hematopoietic progenitor cells.

By virtue of the separability of the system components, the present invention further provides for a composition comprising differentiated cells that are free of both stromal cells and cells expressing a differentiation inducing ligand. In one embodiment of the composition of the instant invention, the differentiated cells are lymphocytes; in a further embodiment, they are T cells.

The present invention further provides for a system for inducing differentiation of a hematopoietic progenitor cell into a T cell, comprising bone marrow derived stromal cell conditioned medium; undifferentiated hematopoietic progenitor cells; and a differentiation inducing ligand attached to a support that induces formation of T cells. One embodiment of the present system provides that the bone marrow derived stromal cell conditioned medium is from OP9 cells. In a further embodiment, the system provides that the undifferentiated hematopoietic progenitor cells are undifferentiated bone marrow derived hematopoietic stem cells. In another embodiment, the system provides that the differentiation inducing ligand attached to a support is Notch ligand.

In an alternate embodiment of the system of the present invention, the bone marrow derived stromal cell conditioned medium is from OP9 cells; the undifferentiated hematopoietic progenitor cells are undifferentiated bone marrow derived hematopoietic stem cells; and the differentiation inducing ligand attached to a support is Notch ligand.

BRIEF DESCRIPTION OF THE DRAWINGS The following drawings form part of the present specification and are included to further demonstrate certain aspects of the present invention. The invention may be better understood by reference to one or more of these drawings in combination with the description of embodiments presented herein.

FIGURE 1 is a diagram showing an illustration of an embodiment of system of the present invention. FIGURE 2 is a diagram showing an illustration of a specific embodiment of a system of the present invention.

FIGURE 3 is a schematic of DLL4-functionalized microbead interacting with hematopoietic progenitor. Streptavidin (Sav) coated microbead is coated with a biotinylated anti-HIS antibody and a polyhistidine tagged DLL4 ligand.

The DLL4 ligand interacts with the Notch receptors on hematopoietic progenitors, triggering Notch signaling and T cell commitment.

FIGURE 4 is a graph showing that microbeads can be efficiently functionalized with Notch ligand DLL4. To assess the efficiency of ligand binding, functionalized beads were stained with anti-DLL4 antibody and FITC anti-rat IgG and analyzed using flow cytometry. Streptavidin coated beads were used as negative controls, (a) Solid histogram represents negative control while unfilled histogram represents functionalized beads. (b) Comparison of coating efficiency for functionalized and streptavidin coated beads. All experiments were performed in triplicate.

DETAILED DESCRIPTION

The present invention generally relates to methods for the ex vivo expansion and use of undifferentiated cells. More specifically, the present invention relates to systems and methods of differentiating undifferentiated cells into a desired linage.

According to the present invention, by providing a paracrine factor from tissue-appropriate stromal cells and a differentiation inducing ligand on a support, an undifferentiated cell may be directed into a more differentiated cell type. The system of the present invention may be used to induce differentiation of any undifferentiated cell that requires cell-derived soluble factors (i.e., paracrine factors) and cell-contact dependent signals. The systems and methods of the present invention are particularly suited to high throughput production of differentiated cells, as well as large scale production of differentiated cells for therapeutic applications. A particular advantage of the present invention is that it does not rely on transfected stromal cells as the signaling entity for the creation of differentiated cells.

The present invention is based in part on a microbead-based, artificial Notch signaling system developed to study stem cell differentiation into the T cell lineage (Figure 1). Magnetic microbeads were functionalized with the Notch ligand DLL4 using streptavidin-biotin binding and antibody-antigen coupling. Immunohistochemistry and flow cytometry analysis indicated about 90% conjugation efficiency. Efficient Notch signaling through these functionalized microbeads was demonstrated through a myotube inhibition assay in C2C12 myoblasts. Thy 1.2⁺ early T cells were successfully generated from mouse bone marrow hematopoietic stem cells (BMHSCs) using DLL4-functionalized beads using both insert and stromal cell (OP9) co-culture conditions, indicating that stem cell-stromal cell physical contact is not necessary for DLL4 directed T cell differentiation. Both insert and mixed co-culture studies with bead to cell ratios of 1:1 generated higher T cell differentiation efficiencies, compared to bead to cell ratios of 5:1. These data demonstrate that this biomaterial-based Notch signaling system elucidates the molecular interactions in T cell development and make ex vivo T cell generation a practical alternative for therapeutic applications. The results for T cell generation indicate that this system can be adapted, through the selection of tissue-appropriate stromal cells and differentiation inducing ligands, to generate other cell types from undifferentiated stem cells.

The systems of the present invention comprise undifferentiated cells, a stromal cell paracrine factor, and a differentiation inducing ligand. Undifferentiated Cells

The undifferentiated cells may be any cells that have not terminally differentiated, for example, stem cells. Suitable undifferentiated cells include those that have the ability to differentiate into specialized cells, and may give rise to one or more lineage-committed cells, which in turn give rise to various types of differentiated cells and tissues. Because undifferentiated cells have the ability to produce differentiated cell types, once differentiated, they may be useful for replacing the function of failing or diseased cells in many tissues and organ systems.

The undifferentiated cells to be differentiated and expanded in the present invention may be isolated from a variety of sources using methods known to one skilled in the art. The undifferentiated cells can be of ectodermal, mesodermal, or endodermal origin. Any undifferentiated cell which can be obtained and maintained in vitro can potentially be used in accordance with the present invention. The undifferentiated cells may be expanded under cell growth conditions, such as conditions that promote proliferation ("mitotic activity") of the cells. In certain embodiments, the undifferentiated cell is a stem cell. Such stem cells include, but are not limited to, hematopoietic stem cells, bone marrow stromal cells (also called mesenchymal stromal cells), which give rise to cell populations that generates bone, cartilage, fat and fibrous connective tissue; stem cells of epithelial tissues such as the skin and the lining of the gut, embryonic heart muscle cells; and neural stem cells.

Negative and positive selection methods known in the art may be used for enrichment of the undifferentiated cells. For example, undifferentiated cells may be sorted based on cell surface antigens using a fluorescence activated cell sorter, or magnetic beads that bind cells with specific cell surface antigens. Negative selection columns can be used to remove cells expressing lineage specific surface antigens.

In certain embodiments, the undifferentiated ceil is a hematopoietic progenitor cell that is capable of differentiating into cells of the T cell lineage. In alternate embodiments, the undifferentiated cell is a hematopoietic progenitor cell that is capable of differentiation into cells of the B cell lineage. Such hematopoietic progenitor cells can be from a newborn mammal, a juvenile mammal, or an adult mammal. Preferred mammals include, for example, humans, non-human primates, mice, pigs, cows, and rats. They can be derived from bone marrow, blood, umbilical cord, stromal cells, fetal tissue, and other sources that are known to those of ordinary skill in the art, or they may be derived from hematopoietic stem cells from such sources. Cells may be obtained from samples from an individual for use in a treatment (e.g., a patient sample) using standard techniques. In a particular embodiment, hematopoietic progenitor cells are obtained from fetal liver tissue, bone marrow, or thymus.

The undifferentiated cells may be cultured in a standard tissue culture dish or a filter insert culture. The concentration of undifferentiated cells in the culture factors known to one skilled in the art. For example, the concentration may depend on the particular undifferentiated cell used, the chosen application, and the volume of cell culture used. In certain embodiments, the undifferentiated cell may be seeded in the culture at density in the range of from about 2.5 x 10³ to about 5 x 10⁵. The culture medium for sustaining the undifferentiated cells in an undifferentiated state can be a conditioned medium (discussed infra) or unconditioned medium. Examples of suitable conditioned medium include Iscove's Modified Dulbecco's Medium (IMDM), DMEM, or αMEM, conditioned from embryonic fibroblast cells (e.g., human embryonic fibroblast cells or mouse embryonic fibroblast cells), or equivalent medium. Examples of suitable non-conditioned medium include, IMDM, DMEM, αMEM, or equivalent medium. The culture medium may comprise serum (e.g., bovine serum, fetal bovine serum, calf bovine serum, horse serum, human serum, or an artificial serum substitute) or it may be serum free. Stromal Cell Paracrine Factors

One or more positive cytokines that promote commitment and differentiation of the undifferentiated cells may also be added to the culture. The cytokines may be human in origin, or may be derived from other species. The concentration of a cytokine in a culture is typically from about 1 to about 100 ng/niL. Suitable cytokines include, but are not limited to, members of the fibroblast growth factor (FGF) family (e.g., FGF-4, FGF-2, Flt-3-ligand, and interleukin-7 (11-7)), and stem cell factor (SCF). The cytokines may be used in combination with equal molar or greater amounts of a glycosaminoglycan such as heparin sulfate. The cytokines are commercially available or can be produced by recombinant DNA techniques and purified to various degrees. Some of the cytokines may be purified from culture media of cell lines by standard biochemical techniques.

The paracrine factors may be present in medium comprising conditioned medium. One of ordinary skill in the art understands that conditioned medium can be collected from cells that were cultured in a medium and the collected medium contains soluble factors secreted by the cells cultured therein. The same media described for culturing undifferentiated cells is appropriate for culturing stromal cells.

The stromal cells provide locally acting soluble factors required for differentiation, i.e., paracrine factors. Any suitable stromal cell may be used, so long as it provides soluble factors required for differentiation of the chosen undifferentiated cell into the desired cell lineage. Examples of suitable stromal cells include, but are not limited to, OP9 cells, STO cells, and S 17 cells. The stromal cell chosen should be based on the desired differentiated cell type from which the undifferentiated cell is differentiated. For example, when the desired differentiated cell type is a T cell, the stromal cells may comprise OP9 cells. When the desired differentiated cell type is a B cell, the stromal cells may also comprise OP9 cells.

Care should be taken in stromal cell culture to minimize spontaneous stromal cell differentiation. The stromal cells can be cultured as a mixed population with the undifferentiated cells or, preferably, as a physically separated population. Cells should be cultured in appropriate density to provide adequate signal for differentiation. In certain embodiments of the systems of the present invention, the stromal cells and the undifferentiated cells may be physically isolated. Such embodiments facilitate isolation of the resultant differentiated cells because, among other things, cell types are not mixed. For example, undifferentiated cells may be cultured in transwell permeable inserts, such as Polyester (PET) Membrane Transwell^®-Clear Inserts commercially available from Corning Life Sciences, which separate the stromal cell layer from the undifferentiated cells.

As used herein, "inserts" refers to permeable supports containing microporous membranes. Such inserts allow cells cultured thereon to uptake and secrete molecules on both their basal and apical surfaces, thereby more closely approximating the in vivo state. Moreover, as described infra, inserts act to physically separate cells of different types, allowing one to easily isolate pure populations of a desired cell type that may be co-cultured with a different cell type. Inserts can be of various materials and pore size; selection of an appropriate insert depends on the cell type and cellular function one is studying.

Other culture methods designed to keep cells separated include, but are not limited to, microfabricated cultures that allows cells to be patterned in a spatially separated manner could be used, and culturing cells in different scaffolds (as in conventional tissue engineering) with one scaffold carrying the supportive cells while the other carrying the undifferentiated cells cultured in close proximity of each other. Differentiation Inducing Ligands and Supports The differentiation inducing ligand provides cell contact-dependent signals required for differentiation. Any suitable differentiation inducing ligand may be used, so long as it provides the contact-dependent signals required for differentiation of the chosen undifferentiated cell. Examples of suitable differentiation inducing ligands include, but are not limited to, Notch ligands, growth factors, major histocompatibility complex (MHC), human leukocyte antigens (HLA), and the like. The differentiation inducing ligand chosen should be based on the desired differentiated cell type from which the undifferentiated cell is differentiated. For example, when the desired differentiated cell type is a T cell, the differentiation inducing ligand may comprise a Notch ligand or an MHC molecule; in Pl 9 embryonal carcinoma cells, RAR and RXR selective ligands induce apoptosis and neuronal differentiation.

A "Notch ligand" is capable of binding to a notch receptor polypeptide present in the membrane of a number of different mammalian cells such as hematopoietic stem cells. Notch ligands may be identified by standard techniques. The Notch receptors that have been identified in human cells include

Notch- 1, Notch-2, Notch-3, and Notch-4. In general, suitable Notch ligands promote and maintain differentiation and proliferation of cells of the T cell lineage.

A Notch ligand may be human in origin, or derived from other species, including mammalian species such as rodent, dog, cat, pig, sheep, cow, goat, and primates.

Particular examples of Notch Ligands include the Delta family. The Delta family includes Delta- 1 (e.g., Dili), Delta-3 (e.g., D113), Delta-like 1, Delta-like 3, Delta-4

(e.g., D114), and Delta-like 4. Notch ligands are commercially available or can be produced by recombinant DNA techniques and purified to various degrees. Notch ligand homologues capable of binding to a Notch receptor, as well as Notch ligand mutants (i.e., a polypeptide having a primary amino acid sequence which differs from the wild type sequence by one or more amino acid additions, substitutions, or deletions) and Notch ligand variants (i.e., a naturally occurring polypeptide which differs from a wild-type sequence) also are suitable for use in the present invention. The differentiation inducing ligand also may be independently provided to the undifferentiated cells on a support, for example, in a manner that allows removal of the differentiation inducing ligand from the resultant differentiated cells. In certain embodiments, the differentiation inducing ligand may be provided by one or more of coating them on a support such as microbead, a scaffold, a liposome, a red blood cell or a red blood cell ghost, and a non-stromal cell. The ligand could also be expressed in a reticulocyte so that it is present on a mature red blood cell or red blood cell ghost. The differentiation inducing ligand may also be provided one or more of coating them on a transwell permeable insert, by encapsulation in degradable microparticles, and transfection of stromal cells that are placed in contact with the undifferentiated cells. Coating differentiation inducing ligands on inserts may be done through physical adsorption or through covalent or non-covalent interactions

(e.g., using neutravidin-biotin binding and antibody-antigen interactions).

Microbeads may be formed by immobilizing a differentiation inducing ligand to a biotinylated microbead through streptavidin-biotin chemistry and antibody-antigen reactions. Because the necessary signals to commit the undifferentiated cells to a desired cell lineage are provided independently, the resulting differentiated cells are substantially pure and, for example, suitable for transplantation or other subsequent use. As used herein, "scaffold" refers to an artificial, biocompatible malleable structure onto which cells are implanted or "seeded," and which can support three-dimensional cell growth, such as tissue or organ growth or regeneration. In addition, scaffolds can be used to deliver biochemical factors, e.g., differentiation inducing ligands, growth factors, cell nutrients, into the body, to support and direct the growth of new cells of an organ or tissue.

Scaffolds can be of natural or synthetic materials, and may be permanent, bioerodable or bioresorbable. Examples of natural scaffold materials include collagen, some linear aliphatic polyesters, chitosan and glycosaminoglycans such as hyaluronic acid. Commonly used synthetic bioerodable scaffold materials include polylactic acid (PLA), polyglycolic acid (PGA); poly (D,L-lactide-co- glycolide) (PGLA) and polycapro lactone (PCL). Scaffolds generally have a high porosity to facilitate cell seeding and diffusion throughout the structure.

Methods

In general, the methods of the present invention comprise providing a cell culture, the culture comprising an undifferentiated cell and a supportive stromal cell, adding a differentiation inducing ligand to the culture, allowing the undifferentiated cell to differentiate into a differentiated cell, and isolating the differentiated cell from the supportive stromal cell and the differentiation inducing ligand. In certain embodiments, the methods of the present invention may be practiced on relatively large numbers of undifferentiated cells in order to produce clinically useful amounts of differentiated cells. Various methods are known in the art for producing such large amounts of undifferentiated cells. For example, undifferentiated cells may be cultured by various known techniques to encourage growth and proliferation, see E. J. Robertson "Teratocarcinomas and Embryonic Stem Cells: A Practical Approach," IRL Press (1987).

The undifferentiated cells should be allowed to differentiate for a sufficient period of time so that the undifferentiated cells form the desired differentiated cells. In general, this period of time may range from about 4 to about 50 days. It will be appreciated that the cells may be maintained for the appropriate amount of time required to achieve a desired result. For example, when the desired differentiated cell is a T cell, to generate a cellular composition comprising primarily immature and inactivated T-cells, the cells may be maintained in culture for about 5 to about 20 days, and cells may be maintained in culture for about another 10 to 15 days, for a total of about 20 to about 30 days, to generate a cellular composition comprising primarily mature T-cells.

In certain embodiments of the methods of the present invention, the undifferentiated cell are hematopoietic stem cells, the supportive stromal cells are OP9 celsl, the differentiation inducing ligand is a DLL4 Notch ligand immobilized to a biotinylated microbead, and the undifferentiated cells differentiates into T cells.

In an alternate embodiment, differentiating an undifferentiated stem cell, wherein the differentiated cell desired is a B cell, requires that undifferentiated cells are hematopoietic stem cells and supportive stromal cells are OP9 cells. In this embodiment, no differentiation inducing ligand is necessary.

In certain embodiments, such a method could be modified and used for tissue engineering organ structures by spatially differentiating single undifferentiated cells into multiple differentiated cell lineages. A typical example would be spatial patterning of the Notch ligand in scaffold structures and directing stem cell differentiation. For example, one can direct undifferentiated cells, for example, BMHSCs, to differentiate using differentiation inducing ligands attached to a support, such as a biodegradable scaffold and supportive stromal cells. The differentiated cells on scaffold areas where no Notch ligand is present will consist of B cells (B cell differentiation being insensitive to Notch signaling), whereas differentiated cells on scaffold areas having Notch ligand will consist of T cells (e.g., as depicted in Figure 2). By spatially patterning a Notch ligand on a polymer scaffold, organized T and B cell structures can be created from a single undifferentiated cell population. In a specific embodiment, the undifferentiated cells are hematopoietic stem cells, the differentiation inducing ligand is a Notch ligand; the scaffold is a PLGA scaffold, and the stromal cells are OP9 cells. Importantly, the method of the present invention does not require cell to cell contact of the stromal cells with the undifferentiated cells. This could provide, among other things, a lymph-node like organoid suitable for use in high-throughput drug and vaccine screening as well for studies of antigen presentation and biology. Composition of Matter

The present invention further provides for a composition of matter comprising undifferentiated cells and a differentiation inducing ligand attached to a support, particularly a scaffold. In particular embodiments, the composition of the present invention can be cultured in situ or in vitro to direct stem cell differentiation and expansion. When in vitro, one may add stromal cell conditioned medium. The undifferentiated cells, support, ligand, and stromal cell conditioned medium are as set forth above with respect to the system of the invention. One embodiment of the present composition of matter provides that the undifferentiated cells are hematopoietic progenitor cells.

In certain embodiments of the methods of the present invention, the undifferentiated cell is a hematopoietic stem cell, the supportive stromal cell is an OP9 cell, the differentiation inducing ligand is a DLL4 Notch ligand immobilized to a biotinylated microbead, and the undifferentiated cell differentiates into a T-cell. Differentiated Cells

The present invention also provides for pure populations of differentiated cells that are free of any other cells, except possible undifferentiated cells. Differentiated cells are any cells that are specialized for a particular function and lack the ability to generate other kinds of cells or revert back to a less specialized cell. In specific embodiments, the differentiated cell is a T cell. Other differentiated cells include, for example, muscle cells and neurons. The differentiated cells may then be used to regrow or regenerate dying or diseased cells in tissue and organs.

Following the methods and systems of the invention, pure populations of differentiated cells may be easily prepared by isolating the differentiated cells grown in insert culture, as described supra.

To facilitate a better understanding of the present invention, the following examples of specific embodiments are given. In no way should the following examples be read to limit or define the entire scope of the invention.

EXAMPLES

To test the systems and methods of the present invention, the following systems were assembled (Figures 1 and 2) and the method conducted as described herein. Example 1. Efficient Functionalization of Notch LiRand DLL4 to Microbeads Conjugation of Notch Ligand DLL4 to Microbeads

The Notch ligand Delta-Like Ligand 4 (DLL4) was conjugated to magnetic microbeads similar to previously published methods for microbead functionalization (Maus et al, Nat. Biotechnol, 2002, 20(2): 143-8; Maus et at, Clin. Immunol, 2003, 106(l):16-22; Trickett et al, J. Immunol. Methods, 2002, 262(1- 2):71-83; Trickett and Kwan, J. Immunol. Methods, 2003, 275(l-2):251-5). Briefly, biotinylated antibodies specific for a histidine tag on recombinant DLL4 were bound to streptavidin coated superparamagnetic polystyrene microbeads. Biotin Binder Kit microbeads (Dynalbiotech, Brown Deer, WI) were washed and incubated with anti-6x HIS tag antibody (R&D Systems, Minneapolis, MN) at 1 μg/ml for 30min at room temperature. After incubation, beads were again washed and further incubated with the Notch ligand DLL4 (R&D Systems, Minneapolis, MN) at 2-4 μg/ml for 30 min at room temperature. Beads were washed and stored at 4⁰C for future use. Immunohistochemistry and flow cytometry analysis of microbeads

Conjugation of DLL4 was confirmed by visualization through fluorescently labeled antibody staining and flow cytometry analysis. Following anti- 6x HIS tag antibody binding and HIS tagged DLL4 loading, microbeads were blocked for 1 hr at 4⁰C with 3% bovine serum albumin (Sigma- Aldrich, St. Louis, MO). Microbeads were then stained with anti-mouse DLL4 antibody (R&D Systems, Minneapolis, MN) and FITC Anti-Rat IgG (eBioscience, San Diego, CA) for 1 hr each at 4⁰C with adequate washing. Rabbit FITC Anti-Rat IgG (Sigma- Aldrich, St. Louis, MO) labeled microbeads served as controls. FACSCalibur (Becton Dickinson, San Diego, CA) and CellQuest 3.1 software (BD Biosciences, San Jose, CA) were used for data acquisition and analysis.

In past studies, Notch ligands such as DLLl and DLL4 have been presented as retrovirally transfected cell-surface ligands in lymphoid differentiation applications, making ligand quantification and scale-up difficult. To address this quantitative hurdle and to develop a biomaterial-based artificial Notch-signaling system, we evaluated whether the Notch ligand DLL4, immobilized on the surface of magnetic microbeads can function in T cell differentiation similar to when presented on the surface of stromal cells. DLL4 was attached to super paramagnetic microspheres using conventional biotin-streptavidin chemistry and antibody-antigen binding. Our goal in the functionalization process was to (a) present the ligand in an efficient manner and (b) preserve the conformational structure of the ligand during the process. To achieve this goal, we first conjugated biotinylated anti-HIS antibody to streptavidin coated beads. Next, we further coated the beads with a recombinant HIS tagged DLL4 Notch ligand. Figure 3 shows the schematic structure of the DLL4- functionalized microbeads.

Results

We assessed the efficiency of the ligand functionalization process by flow cytometry following antibody staining. DLL4-functionalized beads were stained with anti-DLL4 antibody and a FITC anti-rat IgG for ligand detection. Streptavidin coated beads were used as controls. As shown in Figure 4, 90% ± 2.96% of the beads were functionalized with DLL4 whereas control beads showed less than 3% ± 0.97% staining, most likely due to nonspecific binding.

Example 2. DLL4-functionalized Microbeads Can Provide Efficient Notch Signaling

C2C12 myotube inhibition assay for Notch signaling

Notch signaling has been extensively shown to inhibit the formation of myotubes in C2C12 cells. Thus, myotube inhibition has been used extensively as an assay to demonstrate efficient Notch signaling (Luo et al., MoI. Cell. Biol, 1997, 17(10):6057-67; Lindsell et al., Cell 1995, 80(6):909-17; Kopan et al., Development, 1994, 120(9):2385-96; Varnum-Finney et al, J. CeIl ScU, 2000, 113 Pt 23:4313-8). Myoblast differentiation was conducted similar to the methods described by Varnum- Finney et al, supra. Briefly, DLL4 coated beads were cultured in the presence of C2C12 myoblasts and the degree of myotube inhibition was assessed using phase contrast microscopy. C2C12 cells (ATCC, Manassas, VA) were cultured and maintained in DMEM (Invitrogen, Carlsbad, California) with 10% fetal bovine serum (Hyclone, Logan, UT) and 2mM 1-glutamine (Invitrogen, Carlsbad, California). Cells were trypsinized and replated into coated 96 well tissue culture plates at 2x10⁴ cells/ml in differentiation medium (10% equine serum (Hyclone, Logan, UT), 2mM 1- glutamine (Invitrogen, Carlsbad, California), and DMEM (Invitrogen, Carlsbad, California)). Ligand coated beads were then added to wells at various bead to cell ratios of 1 : 1 and 5:1. Uncoated Biotin Binder kit beads (Dynalbiotech, Brown Deer, WI), added at the same bead to cell ratios, served as controls. AU conditions were performed in triplicate. All media was supplemented with the antibiotics penicillin and streptomycin. Cells were incubated at 37⁰C with 5% CO₂ for 5-6 days. Qualitative analysis of myoblast differentiation and Notch activation was conducted using visualization of myofibril formation, as described before (Varnum-Fiπney et al., supra; Rowley et al, Biomaterials, 1999, 20(l):45-53). Phase contrast images using the 2Ox objective of the Leica fluorescent IRB inverted microscope (Meyer

Instruments, San Antonio, TX) were obtained on Day 5-6 to visualize cell fusion and myofibril formation.

The inhibition of myotube formation is a classical assay performed to demonstrate the efficient signaling through Notch ligands. It has been demonstrated that the presence of immobilized Notch ligands inhibits the efficient spontaneous differentiation of C2C12 myoblast cells in myotubes. The degree of myotube inhibition in the presence of DLL4-functionalized microbeads was qualitatively evaluated through phase contrast microscopy after 6 days of culture. Streptavidin coated beads were used as negative controls. The presence of Notch ligand on the microbead surface significantly inhibits myotube length and maturation (data not shown). Although the presence of some myotube formation is suggested by the elongated morphology of the myoblasts, the mature myotube morphology is markedly reduced. Increasing the functionalized bead to cell ratio from 1:1 to 5:1 had no visible effect in terms of increased myotube inhibition or retention of myoblast morphology, as evident from an observed saturation effect. This data suggests a bead to cell ratio of 1:1 is sufficient to transmit Notch signals and significantly inhibit myotube formation in the C2C12 myoblast system.

Example 3. DLL4-functionalized Microbeads Direct Bone Marrow Hematopoietic Stem Cell (BMHSC) to T Cell Lineage in OP9 Co-culture Systems

BMHSC isolation and culture for in vitro T cell development

Stem cells were cultured with DLL4 functionalzied microbeads, unmodified OP9 cells and exogenously added growth factors similar to methods described by Hozumi et al (J Immunol, 2003, 170(10):4973-9). OP9 cells (gift from Tammy Reid, Toronto, Canada; also ATCC No. CRL-2749) were maintained in 20% FBS (Hy clone, Logan, UT), 2.2 g/1 sodium bicarbonate (Invitrogen, Carlsbad, California) and alpha MEM (Invitrogen, Carlsbad, California). They were seeded at the appropriate cell density in 24 well plates one day prior to stem cell seeding to achieve 60% confluence at the time of stem cell seeding. BMHSCs were isolated from 5 week old female C57BL/6 mice (Jackson Laboratory, Bar Harbor, Maine) using standard femur removal protocols. Lin-Ckit+Scal+ HSCs were isolated using magnetic separation (Miltenyi Biotec, Auburn, CA and Dynalbiotech Brown Deer, WI) and seeded at a density of 2 x 10³ cells/well either directly on top of the OP9 cell layer (mixed co-culture) or in Transwell™ permeable inserts (Corning, Acton, MA) (insert co-culture) to assess if stem cell-stromal cell physical contact is absolutely necessary for T cell generation or whether paracrine signaling is sufficient.

Stem cell factor (SCF) (Peprotech, Rocky Hill, NJ) and interleukin-7 (IL-7) (Peprotech, Rocky Hill, NJ) were added to the culture at 50ng/ml and lOng/ml, respectively. DLL4-functionalized or non-functionalized Biotin Binder Kit beads (Dynalbiotech, Brown Deer, WI) were added at defined concentrations to sample and control wells, respectively. AU beads were washed prior to addition. For the mixed co-culture condition, cells were disrupted on Day 4 and single cell suspensions were filtered through a 40μm filter to remove OP9 cells, which generally form aggregates. Cell suspensions were again seeded on fresh monolayers of OP9 cells for continued culture. Beads, growth factors and medium were replenished. For insert culture, Transwell inserts were removed on Day 4 and placed in wells containing fresh OP9 monolayers and replenished with beads, growth factors and medium. On Day 8 cells were trypsinized and separated from beads using DNAse (the streptavidin on these beads are conjugated using a DNA linker) as per manufacturer's protocol (Dynalbiotech, Brown Deer, WI). Flow cytometry

Flow cytometry was performed similar to methods described by us in Liu et al. (Tissue Eng., 2005, l l(l-2):319-30). Cells were washed in FACS buffer (1% bovine serum albumin (Sigma- Aldrich, St. Louis, MO) and 0.05% sodium azide (Sigma-Aldrich, St. Louis, MO) in PBS) twice before staining and blocked for nonspecific binding using anti-mouse CD16/CD32 Fc Block (BD Pharmingen, San Diego, CA) for 10 min at 4⁰C. Staining with antibodies against stage specific T cell markers was performed at concentrations of lμg/lOOμl of FACS buffer and washed prior to staining with secondary fluorescently labeled antibodies, if necessary. Isotype controls (eBioscience, San Diego, CA) were used as negative controls. AU cells were washed in FACS buffer twice and suspended in fresh buffer for analysis. FACSCalibur (Becton Dickinson, San Diego, CA) and CellQuest 3.1 software (BD Biosciences, San Jose, CA) were used for data acquisition and analysis. The following antibodies and fluorochromes were used to assess stage-specific T cell development from BMHSCs: biotinylated CD 19 (Sigma- Aldrich, St. Louis, MO), Thyl.2-FITC, and streptavidin-PE. CD19 is a B cell-specific surface marker while Thy 1.2 has been extensively used as an early T cell marker in differentiation studies (Hozumi et al, Nat. Immunol., 2004, 5(6):638-44; Hozumi et al, J. Immunol., 2003, 170(10):4973-9; Sclimitt et al, Nat. Immunol, 2004, 5(4):410-7; Schmitt and Zuniga- Pflucker, Immunity, 2002, 17(6):749-56). All antibodies were obtained from BD Pharmingen (San Diego, CA) or eBioscience (San Diego, CA) unless otherwise noted.

Results To further demonstrate the functionality of the DLL4-functionalized microbeads, we added these beads to the well-established OP9 co-culture system and examined the effect on T cell differentiation. The OP9 co-culture system has been widely used for lymphoid differentiation applications due to the secretion of lymphoid specific growth factors from OP9 stromal cells (Hozumi et al, 2004, supra; Hozumi et al, 2003, supra; Schmitt et al, 2004, supra; Schmitt and Zuniga-Pfiucker, 2002, supra; Schmitt et al, J. Exp. Med., 2004, 200(4) :469-79; De Smedt et al, Blood Cells MoI. Dis, 2004, 33(3):227-32; Lehar et al, Blood, 2005, 105(4): 1440-7; Nakano, Semin. Immunol, 1995, 7(3):197-203; Nakano et al, Science, 1994, 265(5175):1098- 101; Ciofani et al, J. Immunol, 2004, 172(9):5230-9). However co-culture of stem cells with OP9 cells without the presence of any Notch signaling exclusively generates B cells. Recently, several groups have transfected both the original stem cell source and supportive OP9 cells with various Notch ligands and observed the emergence of cells of the T cell lineage (Hozumi et al, 2004, supra; Hozumi et al, 2003, supra; Sclimitt et al, 2004, supra; Schmitt and Zuniga-Pfiucker, 2002, supra; Schmitt et al, 2004, supra; De Smedt et al, 2004, supra; Lehar et al, 2005, supra). We added the DLL4-functionalized microbeads to the OP9 co-culture system and analyzed the cultures for the expression of B and T cell specific surface markers, CD 19 and Thy 1.2, respectively. We also examined the effect of cell-cell contact on T cell differentiation using the microbead and OP9 system. Transwell™ permeable inserts were used to prevent cell-cell contact with the supportive OP9 layer on the culture well and BMHSCs on insert. Insert culture was compared to conditions where the BMHSCs were seeded directly on the OP9 monolayer. Bead to cell ratio of 1:1 was used for both conditions and streptavidin-coated beads (with no DLL4 attached) were used as negative controls. Cultures with DLL4-functionalized beads gave rise to both Thyl.2⁺ cells and CD19⁺ cells, whereas streptavidin-coated bead culture resulted in CD19⁺ cells only. This held true for both insert and co-culture conditions. In comparison to co-culture conditions, there was a larger percentage of Thy 1.2 cells in insert culture conditions (59.7% vs. 33.4% from co-culture conditions; data not shown). In both insert and co-culture systems, functionalized bead conditions resulted in the commitment of both T and B cells. The sizable B cell population is most likely due to the nonuniform distribution of the functionalized beads and the inability of some BMHSCs to "see" the Notch signal. B cell differentiation efficiencies using streptavidin-coated control beads was comparable for both insert and co-culture conditions (79.9% vs. 73.8%, respectively; data not shown). During the one week differentiation process the original stem cell population underwent a proliferation of at least 2.0 fold for all conditions assayed (data not shown).

Example 4. Defined Ratios of DLL4-Functionalized Microbeads Can Be Used to Commit BMHSCs to the T Cell Lineage

To further investigate the quantitative effects of the DLL4 Notch ligand on T cell commitment, Lin- cKit+ sca-l+ HSCs were cultured with DLL4- coated and uncoated beads in insert cultures and co-culture conditions in the presence of IL-7 and SCF. Cells were harvested at Day 8 and analyzed for CD19 and Thy 1.2 expression using flow cytometry. DLL4-functionalized microbeads were added at various concentrations and the lymphocyte specific-marker expression after 1 week of culture in both insert and co-culture systems was observed, similar to the procedure outlined above. Bead to cell ratios of 1:1 and 5:1 were used for both insert and co- culture conditions.

Results

DLL4-functionalized bead to cell ratios of 1 : 1 resulted in a significant amount of Thy 1.2 expression for both insert and co-culture conditions, whereas a bead to cell ratio of 5:1 gave rise to little if any Thy 1.2 expression (data not shown). Insert culture again gave rise to a higher percentage of Thy 1.2+ cells as compared to the co-culture conditions when cultured with a functionalized bead to cell ratio of 1 : 1 (51.1% vs. 14.35%). The 5:1 bead to cell ratio only gave rise to 2.6% Thyl.2+ cells in insert culture and failed to yield any Thy 1.2+ cells in the mixed condition (data not shown). Low T cell differentiation efficiencies observed in the 5:1 functionalized bead to cell ratio condition may be a result of the high bead concentration in culture. It is likely that the beads have an inhibitory effect of proliferation of progenitors at high densities.

As seen from the statistical summary (Table 1), DLL4 Notch ligand microbeads at a cell ration of 1 :1 was sufficient to drive the differentiation of BMHSCs to the T cell lineage with the expression of Thy 1.2 as shown with immunophenotype analysis.

TABLE 1.

Microbeads: Cells Gated Events Total Events Label Events % Gated % Total

1:1 2000 7812 LR 227 11.35 2.91

1 :1 2000 4907 LR 631 31.55 12.86

5:1 2000 7898 LR 185 9.25 2.34

5:1 2000 7480 LR 290 14.5 3.88

Isotype Controls

FLl (thy 1.2) 1500 1601 LR 1 0.07 0.06

FL2 (CD 19) 1500 1630 LR 25 1.67 1.53

Discussion

Recent studies have indicated a significant role for the Notch ligand DLL4 in T cell commitment and development. Hozumi et al. demonstrated the presence of Thy 1.2⁺ cells from fetal liver hematopoietic progenitors using OP9 cells stably transfected to express DLL4 (Hozumi et al, 2004, supra). These differentiated cells were also shown to undergo VDJ recombination and express CD25 and/or CD44, indicating the maturation to the double negative 3 (DN3) stage of T cell development. Freitas et al. conducted a recent study where the over expression of DLLl or DLL4 in hematopoietic cells was shown to result in the induction of mature single positive (SP) T cells in athymic mice (Hozumi et al, 2004, supra; de La Coste et al, J. Immunol, 2005, 174(5):2730-7; de La Coste and Freitas, Immunol Lett., 2006, 102(l):l-9).

These studies, however, involve mixed co-culture with genetically modified cells and hence suffer from two fundamental drawbacks: (a) it is difficult to quantitatively study the effects of ligand-cell ratio and ligand interaction duration and (b) ultimately high throughput scale up of such mixed co-culture process for large scale production of therapeutic cells from stem cells would be challenging.

The efficient generation of T cells has been achieved through several stem cell-stroma based co-culture systems (Sclimitt et al, Nat. Immunol, 2004, 5(4):410-7; La Motte-Mohs et al, Blood, 2005, 105(4):l431-9; De Smedt et al., Blood Cells MoI Dis., 2004, 33(3):227-32; Poznansky et al, Nat. BiotechnoL, 2000, 18(7):729-34). However limited studies have been reported in generating T cells using biomaterial-based concepts. Poznansky et al, supra, utilized tantalum coated matrices seeded with mouse thymic stroma to provide the necessary molecular cues to commit human hematopoietic stem cells to the T cell lineage. Despite the high efficiency of CD3⁺ T cell generation, the inherent design of this system hinders the isolation of a pure population of T cells, making the large scale generation of T cells difficult.

The OP9-DLL1 Notch signaling, first reported by Zuniga-Pflucker and colleagues also utilizes a mixed co-culture based design to provide direct cell-cell signaling necessary to commit stem cells to the T cell lineage (Zuniga-Pflucker, Nat. Rev. Immunol., 2004, 4(l):67-72). One fundamental limitation of this system, however, is the dependence of OP9 based systems on transfected cells for Notch signaling. The transfection of stroma cells for different Notch ligands can become cumbersome and interfere with normal genetic expression of the OP9 cell (Lehar et al, Blood, 2005, 105(4):1440-7). In addition, the inherent design of the current co- culture system makes large scale T cell generation difficult, specifically for 3D studies in biomimetic environments. Our design addresses the above mentioned limitations by (a) providing Notch signals through immobilization of DLL4 on microbeads (artificial stromal cells) thereby creating a highly controllable, predesigned thymic microenvironment and (b) by physically separating the OP9 cells from the stem cell population during differentiation, which enable us to obtain a purer population of T cells. The magnetic nature of the bead also facilitates the retrieval of T cells eliminating the difficulties involved with obtaining a pure population. In the present invention we have shown that microbeads functionalized with the Notch ligand DLL4 in combination with stromal cell paracrine factors can be used as artificial stromal cells to trigger Notch signaling in myoblasts and commit BMHSCs to the T cell lineage in both co-culture and insert culture systems in a quantitative manner. This study is one of the first to investigate the role of DLL4 in T cell differentiation using a synthetic, biomaterial-based signaling system. The results demonstrate the promise of a bead based system in studying the roles of Notch ligaiids in lymphocyte development and efficiently generating T cells from progenitor cell population using functionalized biomaterials. Streptavidin coated microbeads were functionalized with a Notch ligand using a biotinylated anti-HIS antibody and a HIS tagged recombinant delta like ligand 4. Such microbeads have been used in T cell culture studies, but only for activation applications (Maus et al, Nat Biotechnol, 2002, 20(2):43-8; Maus et al, Clin. Immunol, 2003, 106(1): 16-22; Trickett, and Kwan, J Immunol. Methods, 2003, 275(l-2):251-5). Our approach utilizes a highly directional binding scheme through streptavidin-biotin binding and antigen-antibody interactions that is performed under mild conditions which does not affect the binding site of the ligand and should ensure appropriate confirmation for efficient Notch signaling. Physical characterization of these beads using flow cytometry indicated about 90% functionalization efficiency. To verify whether DLL4-functionalized microbeads can provide efficient Notch signaling, a myotube inhibition assay was performed. C2C12 myoblast cells spontaneously fuse into myotubes in culture. Activation of the Notch pathway has been shown to significantly inhibit myotube formation in these cells (Varnum-Finney et al, J. CeIl ScL, 2000, 113 Pt 23:4313-8). DLL4-functionalized or unmodified beads at various bead to cell ratios were used. We demonstrate that a bead to cell ratio of 1 :1 is sufficient to inhibit myotube formation. Although some myotube-like structures are still evident the tubes appear to be significantly smaller and less structured compared to cells cultured with uncoated beads.

To demonstrate the functionality of the microbeads in lymphocyte development, BMHSCs were cultured in the presence of functionalized beads and either physically separated OP9 stromal cells (insert cultures) or mixed co-culture conditions. Our results indicate that functionalized beads induced a significant percentage of Thyl.2⁺ cells in both cultures. However, insert cultures appeared to generate a higher percentage of cells. The reduced surface area in the Transwell™ inserts leads to higher stem cell-stem cell interactions (for the same number of cells plated) which could be responsible for the increased differentiation efficiency. CD19⁺ B cell differentiation efficiency was similar amongst both insert and co-culture conditions with or without functionalized beads. The presence of some B cells in the Notch functionalized bead culture again indicates non uniform localization of the DLL4 beads in the wells. Still, the differentiation efficiency of Thyl.2⁺ cells for at least the insert condition is comparable to the results described where the Notch ligand was provided by OP9 cells transfected for DLL4 (Hozuni et al, Nat. Immunol, 2004, 5(ό):638-44). Finally, to investigate the quantitative effects of DLL4 on lymphocyte development, bead to cell ratios of both 1 :1 and 5:1 were added in both insert and co- culture conditions. We found that functionalized bead to cell ratios of 1:1 resulted in a significantly higher Thy 1.2+ cell differentiation efficiency than the 5:1 bead to cell ratio in both insert and co-culture conditions. The reduced differentiation efficiency in the 5:1 functionalized bead to cell ratio may be attributed to an inhibitory effect on progenitor proliferation, similar to what has been observed in T cell activation studies with anti-CD3/anti-CD28 microbeads (Ito et ai, J. Immunother., 2003, 26(3):222-33).

In conclusion, we have invented a synthetic biomaterial-based system that can effectively trigger Notch signaling during lymphocyte development from bone marrow-derived stem cells. The system offers the possibility of being quantitative and tunable, enabling the optimization of T cell generation. The systems of the invention could further progress our understanding of ex vivo T cell generation and may allow us to generate functional T cells from stem cells using biomaterial- based approaches. * * *

While the present invention is susceptible to various modifications and alternative forms, specific exemplary embodiments thereof have been shown in the drawings and are herein described. It should be understood, however, that the description herein of specific embodiments is not intended to limit the invention to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the appended claims.

All patents, patent applications, publications, products descriptions, and protocols, and references cited herein are incorporated by reference for all purposes, and specifically for a referenced method or procedure.

Claims

What is claimed is:

1. A conditioned medium system for inducing differentiation of undifferentiated cells, comprising a stromal cell conditioned medium; undifferentiated cells; and a differentiation inducing ligand attached to a support.

2. The system of claim 1, wherein the conditioned medium contains stromal cells.

3. The system of claim 2, wherein the stromal cells are bone marrow derived stromal cells.

4. The system of claim 2, wherein the stromal cells are a stromal cell line.

5. The system of claim 4, wherein the stromal cell line is OP9.

6. The system of claim 2, wherein the stromal cells are physically separate from the undifferentiated cells.

7. The system of claim 1, wherein the undifferentiated cells are hematopoietic progenitor cells.

8. The system of claim 1 , wherein the differentiation inducing ligand support is a bead.

9. The system of claim 8, wherein the bead is a magnetic bead.

10. The system of claim 1, wherein the differentiation inducing ligand support is a scaffold.

11. The system of claim 1 , wherein the differentiation inducing ligand is attached to the support via a biotin-streptavidin link.

12. The system of claim 1, wherein the differentiation inducing ligand and the stromal cell conditioned medium induce differentiation of a hematopoietic progenitor cell.

13. The system of claim 12, wherein the hematopoietic progenitor cell differentiates into a T cell.

14. The system of claim 12, wherein the differentiation inducing ligand is a Notch ligand.

15. A composition of matter comprising undifferentiated cells; and a differentiation inducing ligand attached to a support.

16. The composition of claim 15, wherein the undifferentiated cell a hematopoietic progenitor cell.

17. The composition of claim 15, wherein the composition is an in vitro composition.

18. A method comprising contacting an undifferentiated cell with (a) a stromal cell paracrine factor and (b) a differentiation inducing ligand attached to a support.

19. The method of claim 18, wherein the contacting is done in vitro.

20. The method of claim 18, wherein the undifferentiated cell is a hematopoietic progenitor cell.

21. The method of claim 18, wherein the stromal cell paracrine factor is free of stromal cells.

22. The method of claim 18, wherein the differentiation inducing ligand support is a bead.

23. The method of claim 18, wherein the differentiation inducing ligand support is a scaffold.

24. A composition comprising differentiated cells that are free of both stromal cells and cells expressing a differentiation inducing ligand.

25. The composition of 24, wherein the differentiated cells are T cells.

26. A system for inducing differentiation of a hematopoietic progenitor cell into a T cell, comprising bone marrow derived stromal cell conditioned medium; undifferentiated hematopoietic progenitor cells; and a differentiation inducing ligand attached to a support that induces formation of T cells.

27. The system of 26, wherein the bone marrow derived stromal cell conditioned medium is from OP9 cells.

28. The system of 26, wherein the undifferentiated hematopoietic progenitor cells are undifferentiated bone marrow derived hematopoietic stem cells.

29. The system of 26, wherein the differentiation inducing ligand attached to a support is Notch ligand. W

30. The system of 26, wherein the bone marrow derived stromal cell conditioned medium is from OP9 cells; the undifferentiated hematopoietic progenitor cells are undifferentiated bone marrow derived hematopoietic stem cells; and the differentiation inducing ligand attached to a support is Notch ligand.