US 20060003446 A1
The present invention provides cell populations that are enriched for mesendoderm and mesoderm, and cell populations that are enriched for endoderm. The cell populations of the invention are useful for generating cells for cell replacement therapy.
1. A method of making a cell population enriched for endoderm cells comprising culturing embryonic stem cells in the absence of serum and the presence of activin and Wnt for about one to about six days, isolating brach+/HNF3β+ cells, a culturing said isolated cells in the absence of serum and the presence of activin and an inhibitor of Wnt signalling for at least about one day.
2. The method of
3. The method of
4. The method of
5. The method of
6. A method of making a cell population enriched for endoderm cells comprising culturing human embryonic stem cells in the presence of serum for about two to about ten days, isolating brach+/HNF3β+ cells, and culturing said brach+/HNF3β+ cells in the absence of serum and the presence of activin and an inhibitor of Wnt signalling for at least about one day.
7. The method of
8. The method of
9. The method of
10. The method of
11. An embryonic stem cell in which a nucleic acid encoding a selectable marker is present in the HNF3β locus.
12. The stem cell of
13. The stem cell of
14. The stem cell of
This application is a continuation-in-part of U.S. application Ser. No. 10/514,759, the disclosure of which is incorporated herein by reference, which is a 371 of PCT/US03/15658 filed May. 19, 2003 and which application claims the benefit of U.S. application Ser. Nos. 60/381,617 filed May. 17, 2002 and 60/444,851 filed Feb. 4, 2003, the disclosures of which are incorporated herein by reference.
This invention was made with government support under Grant Nos. 2RO1 HL 48834-09 and 2RO1 HL 65169-02 awarded by the National Institutes of Health. The government may have certain rights in the invention.
During embryonic development, the tissues of the body are formed from three major cell populations: ectoderm, mesoderm and definitive endoderm. These cell populations, also known as primary germ cell layers, are formed through a process known as gastrulation. Following gastrulation, each primary germ cell layer generates a specific set of cell populations and tissues. Mesoderm gives rise to blood cells, endothelial cells, cardiac and skeletal muscle, and adipocytes. Definitive endoderm generates liver, pancreas and lung. Ectoderm gives rise to the nervous system, skin and adrenal tissues.
The process of tissue development from these germ cell layers involves multiple differentiation steps, reflecting complex molecular changes. With respect to mesoderm and its derivatives, three distinct stages have been defined. The first is the induction of mesoderm from cells within a structure known as the epiblast. The newly formed mesoderm, also known as nascent mesoderm, migrates to different positions that will be sites of future tissue development in the early embryo. This process, known as patterning, entails some molecular changes that are likely reflective of the initial stages of differentiation towards specific tissues. The final stage, known as specification, involves the generation of distinct tissues from the patterned mesodermal subpopulations. Recent studies have provided evidence which suggests that mesoderm is induced in successive waves which represent subpopulations with distinct developmental potential. The mesoderm that is formed first migrates to the extraembryonic region and gives rise to hematopoietic and endothelial cells, whereas the next population migrates anteriorly in the developing embryo and contributes to the heart and cranial mesenchyme. These lineage relationships were defined initially through histological analysis and have been largely confirmed by cell tracing studies. While this segregation of developmental fates is well accepted in the field of developmental biology, to date, there are no available methods of isolating mesoderm and endoderm, prior to commitment to these lineages.
The present invention provides a method for isolating mesoderm and definitive endoderm cell populations. These cell populations are useful to identify agents that affect cell growth and differentiation, to identify genes involved in tissue development, and to generate differentiated cells and tissues for cell replacement therapies.
The present invention provides cell populations that are enriched for mesendoderm and mesoderm cells. Mesendoderm cells are defined herein as cells that express brachyury (brach+) and which, in the presence of differentiation-inducing conditions, are capable of generating mesoderm and mesoderm derivatives including cardiac and skeletal muscle, vascular smooth muscle, endothelium and hematopoietic cells, and also are capable of generating endoderm and endoderm derivatives including liver cells and pancreatic cells. Mesoderm cells are defined herein as cells that are brach+ and which, in the presence of differentiation inducing conditions, are capable of generating cardiac and skeletal muscle, vascular smooth muscle, endothelium and hematopoietic cells, and are not capable of generating endoderm and endoderm derivatives.
The present invention further provides cell populations that are enriched for endoderm cells. Endoderm cells are defined herein as cells that do not express brachyury (brach−) and which, in the presence of differentiation-inducing conditions, are capable of generating lung cells, liver cells and pancreatic cells.
The present invention also provides methods of isolating cell populations enriched for mesendoderm and mesoderm cells, and cell populations enriched for endoderm cells.
In another embodiment, the present invention provides methods of identifying agents that affect the proliferation, differentiation or survival of the cell populations of the invention. A method of identifying genes involved in cell differentiation and development of specific lineages and tissues is also provided.
Antibodies that specifically recognize brach+ cells are also provided. The antibodies are useful, for example, for isolating mesendoderm and mesoderm cell populations.
In another embodiment, the present invention provides a method for generating cells in vitro. Such cells are useful, for example, for cell replacement therapy.
The present invention also provides a transgenic non-human mammal having a genome in which DNA encoding a selectable marker is present in the brachyury locus such that one brachyury allele is inactivated and the selectable marker is expressed in cells in which the brachyury locus is transcribed.
FIGS. 3A-C depict the developmental potential of wild type and GFP-Bry ES cells.
FIGS. 5A-C demonstrate the isolation and characterization of GFP and Flk-1 populations.
FIGS. 22A-C demonstrate the effects of activin on development of EBs in serum-free cultures. A) FACS profile showing GFP expression in day 6 EBs differentiated in the presence of 100 ng/ml of activin. B) Kinetics of GFP induction in cultures containing 100 ng/ml of activin. Open circles are EB differentiated in the presence of activin, closed squares are EBs differentiated in absence of activin. C) RT-PCR expression analysis of indicated genes in day 6 EBs grown in the presence (+activin) or absence (−activin) of activin. Numbers indicate day of EB differentiation.
FIGS. 31A-C depict the expression of GFP, hCD4, Brachyury, and HNF3β in developing EBs induced with serum. FIG. 31 A shows dot plots of GFP versus hCD4 expression determined by flow cytometric analysis.
FIGS. 32A-C depict gene expression in different GFP hCD4 expressing population as well as tissue dissected from the primitive streak of mouse embryos. EBs were differentiated in serum for 3.25 days and sorted by the expression of GFP and hCD4.
FIGS. 34A-C depict ES cells differentiated in serum free conditions with Activin A Wnt3a and/or inhibitors of these pathways. ES cells were cultured in serum free media for two days then trypsinized to obtain a single cell population. These cells were then reaggregated with the indicated conditions. At day six, the EBs were plated on matrigel coated dishes in serum free media.
During embryogenesis, the formation of mesoderm is a critical step in the establishment of a body plan and in the development of multiple organ systems such as blood, endothelium, heart and skeletal muscle. The molecular mechanisms that control mesoderm formation, however, are poorly defined. A model system based upon the differentiation of embryonic stem (ES) cells in culture has been used to study mesodermal-derived populations including hematopoietic, endothelial, cardiac and skeletal muscle and adipocyte lineages. The in vitro model supports the induction and specification of mesoderm, but these differentiation events take place in complex colonies known as embryoid bodies (EBs) generated from ES cells. It would be advantageous to isolate mesoderm cell populations from EBs as they are formed, in order to better understand mesoderm formation and tissue development. However, it has not been possible to isolate these populations by cell sorting using antibodies, because antibodies specific for nascent mesoderm cell populations are not well-defined.
Brachyury (also known as T) is the founding member of a family of transcription factors known as T-box genes and was first identified as a naturally occurring mutation in mice. Papaioannou et al. (1998) Bioessays 20:9-19. Heterozygous mice are viable but have a shorter tail than wild type animals. Homozygous mutants, which die at approximately day 10 p.c., lack a notochord and display defects in the development of posterior mesodermal tissues. Through the analysis of chimeric animals, brachyury has been shown to affect the migratory properties of the mesodermal cells. Wilson et al. (1995) Development 121:877-86. Expression analysis revealed a unique and interesting pattern for brachyury. It is expressed transiently in all cells ingressing through the primitive streak as well as in the nascent and early migrating mesoderm. Wilkinson et al. (1990) Nature 343:657-9; Herrmann et al. (1991) Development 113:913-7. Expression is rapidly downregulated in paraxial, lateral and extraembryonic mesoderm and following regression of the steak, is confined to the tailbud and notochord. Given this pattern, brachyury is considered to be one of the best markers of early mesoderm and is used to track the development of this lineage. Brachyury has been identified in all species analyzed, suggesting that its role in mesoderm development is preserved throughout phylogeny. Papaioannou et al. (1998).
In accordance with the present invention, a selectable marker gene has been recombinantly targeted to the brachyury locus. It has been discovered that, following the initiation of ES cell differentiation, the selectable marker is expressed in a pattern that reflects brachyury expression. The selectable marker has allowed the sorting of brachyury positive (Brach+) cells from EBs, and thereby the isolation and characterization of cell populations that are enriched for mesendoderm and mesoderm cells.
The selectable marker exemplified in accordance with the present invention is the enhanced green fluorescence protein (EGFP or GFP). Other selectable markers that will facilitate cell sorting are known to those of ordinary skill in the art and may be used in the present invention. The cDNA encoding GFP is known in the art (and is commercially available, for example as plasmid pEGFP.C1 from Clontech, Palo Alto, Calif.), and may be targeted to the brachyury locus by constructing targeting vectors (GFP-Bry) by methods known in the art. The vectors are preferably designed to replace approximately two-thirds of the first exon of the brachyury gene with a GFP expression cassette.
Brachyury genes from numerous species, including human and mouse, are known in the art and reviewed, for example, by Smith (1997) Current Opinion in Genetics & Development 7:474-480. The GFP expression cassette preferably contains GFP cDNA and one or more translational stop codons to prevent translation of downstream brachyury exons. The cassette may further contain an exon encoding the SV40 polyadenylation signal sequence to prevent transcription of downstream regions of the brachyury gene.
The vectors are introduced into ES cells by methods known in the art to integrate the GFP-Bry construct by homologous recombination. ES cells may be isolated from blastocysts by methods known in the art and disclosed for example by Evans et al. (1981) Nature 292:154-156, Thomson et al. (1995) Proc. Nat'l. Acad. Sci. USA 92; 7844; U.S. Pat. No. 5,843,780; and Reubinoffet al. (2000) Nature Biotech. 18:399. In a preferred embodiment the ES cells are mouse or human ES cells. Following successful targeting the brachyury start codon becomes the start codon of GFP, resulting in the disruption of the targeted brachyury allele. The resulting cells are designated GFP-Bry ES cells. GFP-Bry ES cells are defined herein as ES cells in which one brachyury allele is inactivated and GFP is expressed under the control of the brachyury regulatory elements.
It has been discovered in accordance with the present invention that GFP-Bry ES cells, in which one brachyury allele is inactivated, are viable and develop and differentiate normally. Further, it has been discovered that GFP expression mirrors endogenous brachyury expression. Accordingly, brach+ cells may be isolated by selecting for cells that express GFP. Cells that express GFP may conveniently be isolated by flow cytometry, for example by fluorescence-activated cell sorting (FACS). Methods for sorting cells based on fluorescent properties are well-known to those of ordinary skill in the art.
The present invention also provides GFP-Bry ES cells that have a selectable marker gene recombinantly targeted to the HNF3β locus. HNF3β is known to be expressed in most endodermal cell types. The marker allows the sorting of HNF3β+ cells. The marker exemplified in accordance with the present invention is a truncated human CD4 that lacks most of the intracellular domain. This cell surface molecule is preferred because it cannot transduce signals and antibodies are available for flow cytometry. Other markers that will facilitate cell sorting are known to those of ordinary skill in the art and may be used in the present invention. The cDNA encoding human DC4 is known in the art and may be targeted to the HNF3β locus by methods known in the art. A targeting vector having cDNA encoding the truncated human CD4 cloned between two arms of homology of HNF3β may be electroporated into GFP-Bry ES cells to provide CD4-HNF/GFP-Bry ES cells. The expression of human CD4 in these cells reflects endogenous expression of HNF3β.
Cell populations that are enriched for mesendoderm and mesoderm cells, as defined hereinabove, may be obtained by culturing GFP-Bry ES cells in the presence of serum for a time sufficient to obtain GFP+ cells, for example for from about one to about four days for mouse cells, and sorting and isolating GFP+ cells, for example by flow cytometry. The cell population that is isolated contains at least about 50%, and preferably at least about 75%, and more preferably at least about 90%, and most preferably at least about 95% or at least about 99% mesendoderm and mesoderm cells. The relative amounts of mesendoderm and mesoderm may be varied by adjusting the length of the culture in serum, with shorter culture times favoring the presence of mesendoderm and mesoderm patterned to the hematopoietic and endothelial lineages, and longer culture times favoring the presence of mesoderm patterned to the cardiac and skeletal muscle lineages. For example, a cell population enriched for mesoderm may be obtained by culturing in serum for about 2.5 to 4.5 days, followed by sorting and isolating GFP+ cells. Culturing in the presence of serum is defined herein as culturing in media supplemented with animal serum, for example fetal calf serum (FCS). In a preferred embodiment, the media is supplemented with from about 5% to about 25% serum. The optimal concentration may be serum batch dependent and can be determined by one of ordinary skill in the art.
Cell populations that are enriched for mesendoderm and mesoderm cells may be obtained from GFP-Bry ES cells generated from human ES cells by a similar method in which the length of time of culture in serum is lengthened to account for difference in times of differentiation in vitro for human and mouse cells. Accordingly, GFP-Bry ES cells generated from human ES cells are cultured in serum for a time sufficient to obtain GFP+ cells, for example about 2 to about 18 days, before sorting and isolating GFP+ cells.
For both mouse and human cell populations, it can be easily determined whether the isolated cells have differentiated beyond mesoderm, for example to hemangioblasts, by assaying for the presence of the tyrosine kinase receptor, human KDR or mouse Flk-1. KDR and Flk-1 are not expressed in mesendoderm and nascent mesoderm, but as these cells differentiate to a hemangioblast/pre-erythroid population, KDR or Flk-1 expression is detectable. KDR+ and flk-1+ cells may be identified by flow cytometry using antibodies to KDR or Flk-1. Such antibodies are known in the art, and may also be generated using standard methods of antibody production. The cell populations enriched for mesendoderm and mesoderm may be further enriched by removing KDR+ or Flk-1+ cells by cell sorting.
As depicted in
Cell populations that are enriched for endoderm cells may be obtained from GFP-Bry ES cells generated from human ES cells by culturing the GFP-Bry ES cells in the presence of serum for about 2 to 10 days, and then sorting and isolating GFP+ cells followed by culturing the GFP+ cells in the absence of serum for from about 1 to about 15 days.
The populations enriched for endoderm cells may be further enriched by identifying and sorting out KDR+ or Flk-1+ cells as described above.
It has further been discovered in accordance with the present invention that cell populations enriched for endoderm may be obtained by culturing GFP-Bry embryonic stem cells in the absence of serum and in the presence of the growth factor activin, for about two to about ten days, and isolating cells that express brachyury. The amount of activin is sufficient to induce differentiation of embryonic stem cells to endoderm. Such differentiation may be measured by assaying for the expression of genes associated with endoderm development, including for example HNF3β, Mix1-1, Sox17, Hex-1 or pdx-1. In a preferred embodiment, the concentration of activin is at least about 30 ng/ml. In another preferred embodiment the concentration of activin is about 100 ng/ml. In another embodiment, the GFP-Bry embryonic stem cells are cultured in the absence of serum and the presence of activin and an inhibitor of Wnt signalling for about two to about 10 days, and cells expressing brachyury are isolated. In a preferred embodiment the inhibitor is DKK-1. In other preferred embodiments the concentration of DKK-1 is at least about 30 ng/ml, or about 100 ng/ml.
Cell populations enriched for mesoderm may be obtained by culturing GFP-Bry embryonic stem cells in the absence of serum and the presence of activin for about two to about ten days, and isolating cells that express brachyury. The amount of activin is sufficient to induce differentiation of embryonic stem cells to mesoderm, but insufficient to induce differentiation to endoderm. Differentiation to mesoderm may be measured by assaying for the expression of genes associated with mesoderm development, including for example GATA-1, and the absence of expression of genes associated with endoderm development. In a preferred embodiment, the concentration of activin is less than 30 ng/ml. In another preferred embodiment the concentration of activin is about 3 ng/ml.
It has been further discovered in accordance with the present invention that cell populations enriched for endoderm may be obtained by culturing embryonic stem cells in the absence of serum and the presence of activin and a Wnt molecule for about one to about six days, isolating brach+/HFN3β3+ cells, and culturing the isolated cells in the absence of serum and the presence of an inhibitor of Wnt signaling for at least about one day. Isolation of brach+/HFN3β3+ cells may be facilitated by using the C4-HNF/GFP-Bry ES cells described hereinabove and selecting for CD4 and GFP, for example by cell sorting.
Wnt refers to a family of polypeptides well-known in the art. See, e.g. U.S. Pat. No. 6,159,462. The Wnt growth factor family includes proteins encoded by at least ten genes in mice and at least seven genes in human. In a preferred embodiment of the present invention, the Wnt molecule is recombinant Wnt 3d.
Inhibitors of Wnt signalling are also well-known in the art and include, for example, Wnt-I disclosed in U.S. Pat. No. 6,844,422 and Dickkopf-1 (DKK-1) disclosed by Glinka et al. (1998) Nature 391:357-362.
The present invention also provides a method of making cell populations enriched for endoderm comprising culturing mouse embryonic stem cells in the presence of serum for about two to about four days, or human embryonic cells in the presence of serum for about two to about ten days, sorting and isolating brach+/HFN3β3+ cells, and culturing the isolated cells for about one to about ten days in the absence of serum and the presence of activin and an inhibitor of Wnt signalling. Isolation of brach+/HFN3β3+ cells may be facilitated by using the CD4-HNF/GFP-Bry ES cells as described above.
In each of the foregoing two methods, the amounts of activin and Wnt inhibitor are sufficient to induce differentiation of ES cells to endoderm. Differentiation may be measured by assaying for the expression of genes associated with endoderm development. In a preferred embodiment the concentration of activin is at least about 30 ng/ml. In another preferred embodiment the concentration of activin is about 100 ng/ml. In a preferred embodiment the Wnt inhibitor is DKK-1 and the concentration of DKK-1 is at least about 30 ng/ml. In another preferred embodiment the concentration of DKK-1 is about 100 ng/ml.
The present invention further provides a method of identifying agents that affect the proliferation, differentiation or survival of the cell populations described above. The method comprises culturing cells from one of the cell populations described hereinabove in the absence and presence of an agent to be tested, and determining whether the agent has an effect on proliferation, differentiation or survival of the cell population. The agent to be tested may be natural or synthetic, one compound or a mixture, a small molecule or polymer including polypeptides, polysaccharides, polynucleutides and the like, an antibody or fragment thereof, a compound from a library of natural or synthetic compounds, a compound obtained from rational drug design, or any agent the effect of which on the cell population may be assessed using assays known in the art, for example standard proliferation and differentiation assays as described in U.S. Pat. No. 6,110,739. Such agents are useful for the control of cell growth and differentiation in vivo and in vitro.
The present invention further provides a method of identifying genes involved in cell differentiation and development of specific lineages and tissues. The method comprises isolating populations of GFP+ cells of the invention after different amounts of time in culture, comparing gene expression profiles in the different populations, and identifying genes that are uniquely expressed in a population. In a preferred embodiment, microarray analysis and subtractive hybridization are used to compare gene expression profiles.
In another embodiment, the present invention provides methods of making antibodies that recognize brachyury positive (brach+) cells but not brachyury negative (brach−) cells. Polyclonal antibodies may be made by injecting an animal with the cells of the invention in an immunogenic form. Also, antibodies may be made by identifying cells surface markers present in GFP+ but not GFP− cells, and making antibodies against the markers or fragments thereof. The antibodies may be monoclonal or polyclonal, and may be fragments, genetically engineered antibodies, single chain antibodies, and so on. Antibodies may be made by methods well-known in the art. Such antibodies are useful for identifying and isolating brach+ cells such as mesendoderm and mesoderm.
The present invention also provides a method for generating mammalian cells in vitro. In one embodiment, the method comprises culturing cells from a cell population enriched in mesendoderm and mesoderm cells under conditions effective for the differentiation of mesoderm into cardiac muscle, vascular smooth muscle, endothelium or hematopoietic cells. Conditions effective for differentiation into the various cell types in vitro are known in the art. In another embodiment, the method comprises culturing cells from a cell population enriched in endoderm cells under conditions effective for the differentiation of endoderm into liver cells or pancreatic cells. Effective conditions for such differentiation are known in the art. The production of insulin-producing pancreatic islet cells is specifically contemplated.
As demonstrated in accordance with the present invention, brach+ cells isolated from different aged EBs have different developmental potentials. Brach+/Flk− cells from about day 3 mouse EBs efficiently generate hemotopoietic and endothelial lineages, while the cells from about day 3 to 10 EBs generate cells of cardiomyocyte lineages. Accordingly, by adjusting the time of culture of the ES cells used for obtaining the cell population enriched for mesendoderm and mesoderm, one of ordinary skill in the art can select for efficient production of hemotopoetic and endothelial lineages or cardiomyocyte lineages.
Such cells are useful, for example, for cell replacement therapy for the treatment of disorders that result from destruction or dysfunction of a limited number of cell types. Such disorders include diabetes mellitus, liver failure, heart failure, cardiovascular and other vascular disease, Duchenne's muscular dystrophy, osteogenesis imperfecta, and disorders treatable by bone marrow transplant, for example leukemias and anemias. See, Odorico et al., (2001) Stem Cells 19:193-204.
The cell populations of the present invention are useful for generating differentiated cells and tissues for cell replacement therapies. The suitability of the cell populations of the present invention for cell replacement therapy may be assessed by transplanting the cells into animal models of disorders that are associated with the destruction or dysfunction of a limited number of cell types. For example, the fumarylacetoacetate (FAH) deficient mouse disclosed for example by Grompe et al. (1993) Genes & Dev. 7:2298, incorporated herein by reference, provides a model for liver failure. FAH deficient mice suffer from progressive liver failure and renal tubular damage unless treated with NTBC (2-(2-nitro-4-trifluoromethyl benzoyl)-1,3-cyclohexedione) or transplanted with normal hepatocytes. These mice thus provide an ideal model for testing the potential of cells with characteristics of immature hepatocytes generated from EBs. Methods for transplantation of hepatocytes into FAH deficient mice removed from NTBC are known in the art and disclosed for example by Oversturf et al. (1996) Nature Genet. 12:266-273. Normal liver function is indicated by survival of the mice, and may also be assessed by measuring serum aspartate transaminase levels, plasma bilirubin levels, and by determining normal structure of the regenerated liver.
Animal models for other disorders that result from the destruction or dysfunction of particular cells types are known in the art. Such models may similarly be used to assess other cell populations of the present invention.
The present invention also provides a transgenic non-human mammal in which DNA encoding a selectable marker is present in the brachyury locus such that one brachyury allele is inactivated and the selectable marker is expressed in cells in which the brachyury locus is transcribed. In a preferred embodiment the mammal is a mouse and the selectable marker is GFP. In particular, the transgenic mouse has a genome comprising a transgene in which a DNA sequence encoding GFP is operably linked to brachyury regulatory elements, and the transgene is expressed in cells that normally express brachyury. The transgenic mouse may be obtained by injecting the GFP-Bry ES cells described hereinabove into blastocysts, which are then implanted into pseudopregnant females. Transgenic pups are identified by the short-tail phenotype associated with brach +/−, and by molecular analysis. Such transgenic animals are useful for obtaining early embryos from which to isolate mesoderm to be used in accordance with the methods of the invention, and for the identification, isolation and characterization of any adult cell populations that express the brachyury gene. Such cells may represent novel stem cell populations.
All references cited herein are incorporated herein in their entirety.
The following examples serve to further illustrate the present invention.
ES cell growth and differentiation. ES cells were maintained on irradiated embryonic feeder cells in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 15% fetal calf serum (FCS), penicillin, streptomycin, LIF (1% conditioned medium) and 1.5×10−4 M monothioglycerol (MTG; Sigma). Two days prior to the onset of differentiation, cells were transferred on gelatinized plates in the same media. For the generation of EBs, ES cells were trypsinized and plated at various densities in differentiation cultures. Differentiation of EBs was carried out in 60 mm petri grade dishes in IMDM supplemented with 15% FCS, 2mM L-glutamine (Gibco/BRL), transferrin (200 ug/ml), 0.5mM ascorbic acid (Sigma), and 4.5×10−4 M MTG. Cultures were maintained in a humidified chamber in a 5% CO2/air mixture at 37° C.
Serum Free Medium. Two different serum-free media were used in different aspects of the following examples: IIMD supplemented with Knockout SR (Gibco BRL) and StemPro 34 (Gibco BRL).
Methylcellulose Colony Assay. A) Blast colonies: For the generation of blast cell colonies (BL-CFC assay), EB-derived cells were plated at 0.5×1.5×105 cells/ml in 1% methylcellulose supplemented with 10% FCS (Hyclone), vascular endothelial growth factor (VEGF; 5 ng/ml), c-kit ligand (KL; 1% conditioned medium), IL-6 (5ng/ml) and 25% D4T endothelial cell conditioned medium (Kennedy et al. (1997) Nature 386:488-93). Transitional colonies were generated in the absence of VEGF. Colonies were scored following four days of culture. B) Hematopoietic colonies: For the growth of primitive and definitive hematopoietic colonies, cells were plated in 1% methylcellulose containing 10% plasma-derived serum (PDS; Antech), 5% protein-free hybridoma medium (PFHM-II; Gibco-BRL) plus the following cytokines: c-kit ligand (KL; 1% conditioned medium), erythropoietin (2 U/ml), IL-11 (25 ng/ml), IL-3 (1% conditioned medium), GM-CSF (3 ng/ml), G-CSF (30 ng/ml), M-CSF (5 ng/ml), IL-6 (5 ng/ml) and thrombopoietin (TPO; 5 ng/ml). Cultures were maintained at 37° C., 5% C02. Primitive erythroid colonies were scored at day 5-6 of culture, whereas definitive erythroid (BFU-E), macrophage, and multilineage colonies were counted at 7-10 days of culture. C-kit ligand was derived from media conditioned by CHO cells transfected with KL expression vector (kindly provided by Genetics Institute). IL-3 was obtained from medium conditioned by X63 AG8-653 myeloma cells transfected with a vector expressing IL-3. VEGF, GM-CSF, M-CSF, IL-6 ,IL-11, activin BMP2, BMP4, bFGF, FGF8, and 1 hh were purchased from R&D systems.
Reaggregation Cultures. Cells were cultured at 2×105 per ml IMDM supplemented with 15% FCS (or Knockout SR), 2mM L-glutamine (Gibco/BRL), 0.5 mM ascorbic acid (Sigma), and 4.5×104 M MTG in 24-well petri-grade plates. These were used to prevent adhesion of the cells to the bottom of the well.
Cardiac muscle assays. GFP+ cells were reaggregated in IMDM supplemented with 15% serum replacement. Twenty hours later the aggregates were cultured in wells of either a 24- or 96-well plate in IMDM with 10% serum replacement (serum-free). The wells were pre-treated with gelatin. Cultured were monitored daily for the development of the appearance of beating cells. Beating cells were usually detected between days 2 and 6 of culture.
Cell surface markers staining and FACS analysis. Standard conditions were used to stain the cells. Stained suspensions were analyzed on a FACScan (Becton Dickinson, CA).
Gene Expression Analysis For the poly A+ RT-PCR analysis the method of Brady et al. ((1990) Meth. in Mol. and Cell Bio. 2:17-25) was used. Reverse transcription, poly-A tailing and PCR procedures were performed as described, with the exception that the X-dT oligonucleotide was shortened to 5′-GTTAACTCGAGAATTC(T)24-3′. The amplified products from the PCR reaction were separated on agarose gels and transferred to a Zeta-probe GT membrane (Biorad) or transferred to the membrane with a slot blot apparatus (Schleicher & Schuell). The resulting blots were hybridized with 32P randomly primed cDNA fragments (Ready-to-Go Labelling, Pharmacia) corresponding to the 3′ region of the genes (for all except β-H1). A β-H1-specific probe was prepared by annealing two oligonucleotides, (5′-TGGAGTCAAAGAGGGCATCATAGACACATGGG-3′, 5′-CAGTACACTGGCAATCCCATGTG-3′) which share an 8 base homology at their 3′ termini. This β-H1 specific otigonucleotide was labeled with 32P using a Klenow fill-in reaction. For gene specific PCR, total RNA was extracted from each sample with RNeasy mini kit and treated with RNase free DNase (Qiagen). Two microgram of total RNA was reverse-transcribed into cDNA with random hexamer using a Omniscript RT kit (Qiagen). PCR was carried out using appropriate oligonucleotides. The PCR reactions were performed with 2.5 U of Taq polymerase (Promega), PCR buffer, 2.5 mM MgCI2, 0.2 uM of each primer and 0.2 mM dNTP. Cycling conditions were as follows; 94° C. for 5 min followed by 35 cycles of amplification (94° C. denaturation for 1 min, 60° C. annealing for 1 min, 72° C. elongation for 1 min) with a final incubation at 72° C. for 7 min.
Under appropriate conditions in culture, embryonic stem (ES) cells will differentiate and form three dimensional colonies known as embryoid bodies (EBs) that contain developing cell populations from a broad spectrum of lineages. Smith (2001) Annu. Rev. Cell Dev. Biol. 17:435-62. Among these EB-derived populations, one can detect mesodermal derivatives including those of the hematopoietic, endothelial, cardiac muscle and skeletal muscle lineages.
In order to track the onset of mesoderm in EBs and to isolate cells representing this population, the green fluorescence protein (GFP) was targeted to the brachyury locus. The targeting construct contained the GFP cDNA, and artificial intron, SV40 poly(A) sequences and a loxP flanked neo cassette in the first exon and is depicted in
A BAC clone carrying the entire mouse Brachyury (Bry) gene was isolated by PCR screening of a 129/O1a strain genomic library (Genome Systems) with primers 5′-AAGGAGCTAACT AACGAGATGAT-3′ and 5′-TACCTTCAGCACCGGGAACAT3′. These primers anneal within the first and second Bry exon, respectively, and amplify a diagnostic band of ˜600 bp. An approximately 3 kb long PstI restriction fragment carrying the 1 exon of the Bry gene along with more than 2 kb of 5′ flanking region was identified and subcloned from the BAC into plasmid pBSK (Strategene), resulting in construct pBSK.Bry-5′. Approximately 2 kb of the region immediately upstream of the start codon were sequenced to identify appropriate primer annealing sites for the construction of vectors.
Oligos 5′-GCTAGCTAATGGATCCA-3′/5′- GATCTGGATCCA TTAGCTAGCTGCA-3′ and 5′- GATCTTAATGAACGGCAGGTGGGTGCGCGTCCGGA G-3′/5′ TCGACTCCGGACGCGCACCCACCTGCCGTTCATTAA-3′ were inserted into the PstI/SaII sites of plasmid pBSK to create a new, more suitable polylinker with two successive translational stop codons and an artificial 3′ splice site (construct pBry-AA). Plasmid pEGFP.C1 (Clontech) was double-digested with NheI/BgIII and the resulting ˜760 bp DNA fragment encoding EGFP without stop codon was cloned into the NheI/BgIII sites of pBRY-AA, resulting in construct pBry-AB. An XhoI/SaII fragment of plasmid pL2-Neo2 carrying a loxP-flanked neomycin resistance gene was inserted into the SaII site of pBry-AB to give rise to plasmid pBry-AC (transcription of EGFP and Neo in same direction).
A 556 bp XmaI/MluI fragment carrying a consensus splice donor site, an artificial intron, a splice acceptor site and a short exon including the SV40 polyadenylation sequence, was excised from the commercial expression vector pBK-CMV (Stratagene). This fragment was inserted into plasmid pBry-AC in the following way: the XmaI end was ligated into the BspEI site following the last EGFP codon, whereas the Mlu end was inserted along with oligos 5′-CGCGTTACTAGTAAGACGTCT-3′/5′-CCGGAGA CGTCTTACTAGTAA-3′ as linkers into the BspEI site located just upstream of the loxP-neo-loxp cassette. Resulting construct: pBry-AE. An ˜1.9 kb XhoI/SaII fragment encoding the HSV thymidine kinase gene was inserted into the XhoI site of pBry-AE to allow selection against random integration (construct pBry-AH). A NotI/Eco47III fragment encoding the “short arm” of homology was excised from pBry-AF and cloned into the NotI/Eco47III sites of pBry-AH to give rise to plasmid pBry-AI. The “long arm” of homology was excised from pBry-AK with SaII and inserted in the correct orientation into the Sail site of pBry-AI to yield final targeting vector B.
Embryonic stem cells (E14. 1, 129/Ola Hooper et al. (1987) Nature 326:292) were electroporated with NotI-linearized targeting vector pBry-AM. Four dishes with transfected cells were subjected to G418 single-and another four dishes to G418+Gancyclovir (Ganc) double-selection. Clones that had undergone a homologous recombination event were identified by PCR with one primer (5′-CAGGTAGAACCCACAA CTCCGAC-3′) annealing to genomic sequences in the 5′ region of the Bry gene, just upstream of the “short arm of homology”, the other (5′-CCGGACACGCTGAACTTGTGGC-3′) to the 5′ portion of EGFP (diagnostic band: approximately 1.3 kb). Correctly targeted clones were confirmed by Southern blot analysis: genomic DNA of candidate clones was digested with HincII and hybridized to a probe located outside of the targeting construct. The probe was derived from the Bry 5′ flanking region (-2018 to -1249 with respect to the Bry ATG start codon) by PCR using the oiigonucleotide pair 5′-ACAGGATCCCTAAGCCTCAAAGAGTCGCT-3′/5′-TCTTGGATCCTCCTAT CCTATCCCGAAGCTCCT-3′. 384 G418 single- and 80 G418+Ganc double-selected clones were screened, of which 4 respectively 3 proved to be positive, corresponding to a targeting efficiency of 1.04% and 3.75%. Two positive clones were transiently transfected with a modified Cre recombinase expression vector to remove the neo gene. These targeted clones are referred to hereinafter as GFP-Bry ES cells.
Brachyury is expressed transiently in developing EBs with the onset preceding the expression of genes that define the establishment of the hematopoietic and endothelial lineages. To determine whether GFP expression in GFP-Bry ES cells accurately reflects expression of the brachyury gene during EB development, GFP expression was assessed.
A typical expression pattern during a 6-day EB differentiation period is shown in
The possibility that inactivation of a single brachyury allele could have detrimental effects on the in vitro developmental potential of the ES cells was assessed. As indicated, heterozygous mice demonstrate a mild phenotype. To determine if the GFP-Bry ES cells display any detectable defects in hematopoietic development, EBs generated from them were analyzed for hematopoietic precursor and blast colony-forming cell (BL-CFC) content and for gene expression patterns. The data in
To determine if brachyury+ cells could be isolated based on GFP expression, the GFP+ population from day 3.5 EBs was sorted and analyzed for expression of appropriate genes.
This example demonstrates that GFP+ cells can be isolated from day 3.5 EBs by cell sorting. Gene expression analysis of the GFP+ and GFP− fractions shows that brachyury expression segregates predominantly to the positive fraction, a finding which clearly demonstrates that fractionation based on GFP provides a method for isolating brachyury positive cells. In addition to brachyury, the receptor tyrosine Flk-1 involved in early hematopoietic and endothelial development is also expressed at higher levels in the positive than in the negative fraction. In contrast, Rex-1 and Pax-6, markers of early ectoderm and neuroectoderm, are expressed in the GFP− fraction. These findings demonstrate that expression of GFP in the context of brachyury can be used to separate mesoderm from ectoderm.
Flk-1 has been shown to be essential for the establishment of the hematopoietic and endothelial lineages in the early embryo and is expressed on the earliest precursors of these lineages, including the BL-CFC [Faloon et al. (2000) Development 127:1931-41]. Given this pivotal role in blood and vascular development, its expression within the GFP+ population was hypothesized to define a subpopulation of mesoderm undergoing specification to these lineages. To investigate this possibility further, EBs were analyzed at several stages of development for the presence of GFP and Flk-1 positive cells. In the experiment illustrated in
This example demonstrates that the brachyury fraction of day 3 and day 3.5 EBs can be separated into two fractions based on Flk-1 expression: brachyury+/Flk-1− (GFP+/Flk-1−) and brachyury+/Flk-1+ (GFP+/Flk-1+) (
The expression patterns observed in
This example indicates that when isolated and recultured for 20-24 hours, each of the three populations isolated from day 3 EBs continued to differentiate and in a pattern that indicates that these populations represent a developmental continuum. For instance, GFP−/Flk-1− gave rise to GFP+/Flk-1− cells which in turn gave rise to GFP+/Flk-1+. These changes in cell surface characteristics were associated with the expected changes in developmental potential. The GFP+/Flk-1− fraction contained few hematopoietic/endothelial precursors (BL-CFC) prior to culture. Following culture, these precursors were detected, clearly demonstrating that the GFP+/Flk-1− fraction from day 3 EBs does contain the potential to give rise to Flk-1+ cells with hematopoietic and endothelial potential.
The foregoing examples demonstrated that GFP+/Flk-1− cells isolated from day 3.0 EBs efficiently generated GFP+/Flk-1+ cells and BL-CFC following overnight culture. To determine if this pre-BL-CFC potential was specific to the GFP+/Flk-1− fraction isolated at this stage of development, GFP+/Flk-1− cells from different aged EBs were assayed for their ability to give rise to BL-CFC. As shown in
Given the sequence of events in the mouse embryo in which the development of hematopoietic and endothelial mesoderm is followed by the development of cardiac and cranial mesoderm, the cardiomyocyte potential of the isolated populations was determined. For this analysis, aggregates from the cultures of the different populations were transferred to microtiter or 24-well plates in serum-free medium and monitored for the development of beating cell masses, indicative of cardiomyocytes. These conditions are known to support the efficient development of cardiomyocytes from the reaggregated cells. As an independent confirmation of the cardiomyocyte nature of the cells within these masses, a representative group was transferred to microscope slides, fixed and stained for the presence of the cardiac-specific isoform of Troponin T. All beating cell masses analyzed were found to contain Troponin T positive cells indicating that they were cardiomyocytes. Using this assay, the cardiomyocyte potential of the reaggregated GFP+/Flk-1− and GFP+/Flk-1+ fractions from different staged EBs was determined. For comparison, the BL-CFC potential of the freshly isolated GFP+/Flk-1+ cells and of the cultured GFP+/Flk-1− cells was analyzed.
As shown in Table 1, the BL-CFC potential of the different fractions was similar to that observed in previous experiments. The GFP+/Flk-1+ cells isolated from the three different EB populations generated BL-CFC, with the highest number found at day 3.5 and 4.0. The pre-BL-CFC potential in the GFP+/Flk-1− cells was greatest at day 2.75 and decreased significantly at 3.5 and 4.0. The cardiomyocyte potential of the fractions showed a reverse pattern. A significant proportion (>80%) of the transferred aggregates from day 3.5 and 4.0 GFP+/Flk-1− cells generated beating cardiomyocytes. No beating cells were observed in the aggregates generated from the earliest (day 2.75) GFP+/Flk-1− cells. Beating cells were not detected in aggregates generated from the GFP+/Flk-1+ cells isolated from EBs at any of the three stages of development. The findings from this analysis are consistent with the notion that GFP+/Flk-1− cells isolated at different stages have different potentials. Those that develop early appear to have a hemangioblast fate, whereas those that develop later generate the cardiac lineage and possibly other populations. The GFP+/Flk-1+ populations appear to have lost the cardiomyocyte potential and may be restricted to the hemangioblast lineages. Given these findings, the early developing (day 2.75-3.0) GFP+/Flk-1− cells are referred to as pre-hemangioblast mesoderm whereas the population that develops between day 3.5 and 4.0 are referred to as pre-cardiac mesoderm. The day 3.0-3.5 GFP+/Flk-1+ populations generate BL-CFC, whereas those isolated from later stage EBs (day 4.0) contain primitive erythroid progenitors, indicating the onset of hematopoietic commitment. Given this developmental potential, the GFP+/Flk-1+ population is referred to as hemangioblast mesoderm.
Examples 5 and 6 demonstrate that GFP+cells isolated from different aged EBs have different developmental potential. As indicated in the previous examples, GFP+/Flk-1− cells from day 3 EBs efficiently generate both hematopoietic and endothelial lineages. These cells did not give rise to cardiotiyocytes (heart tissue) as demonstrated by the lack of beating cell masses. In contrast, GFP+ cells from day 4 EBs gave rise to few Flk-1+ cells and BL-CFC following culture. This population did, however, generate cells of the cardiomyocte lineage. These findings demonstrate that the GFP+ (brachyury+) fraction isolated from different aged EBs have become patterned to distinct populations with different developmental fates. In addition to the conditions used and potentials observed in the foregoing examples, other potentials may be observed by altering conditions and additives.
To assess the role of serum in the development of brachyury+ cells, EBs were differentiated the absence of serum. EBs did develop under these conditions, although they were somewhat smaller than those found in normal conditions. In the absence of serum, no GFP+ cells were detected within these EBs, indicating that mesoderm was not induced under these conditions (
Studies using model systems such as Xenopus and Zebrafish have suggested that mesoderm and endoderm develop from a common precursor population known as mesendoderm. To determine whether or not a mesendoderm stage of development does exist in EBs, conditions for the development of the endoderm lineage were established. As a first step in establishing culture conditions for the development of this lineage, the amount of serum in the differentiation cultures was varied. EBs were generated in the presence and absence of serum (SP34 and then IMDM plus SR) and assayed at different stages for expression of genes associated with ectoderm, endoderm and mesoderm development. For the ectoderm lineage, development of the neuronal lineage was assessed analyzing expression of PAX6, Wnt1, neruoD and neurofilament (NFL). These genes are known to be expressed at different stages of neruonal development. The early stages of endoderm development were monitored by expression of HNF3β. To evaluate later stages of endoderm development and specification, genes involved in liver development were analyzed. These included Hex, α-fetoprotein (AFP), HNF4, Albumin (Alb), α-1-antittrypsin (AAT) and tyrosine aminotransferase (TAT). Mesoderm development was monitored by expression of brachyury and GATA-1. In addition to gene expression patterns, neuronal development was assayed by monitoring neurite outgrowth from EBs. The neuronal nature of these neurites was demonstrated by βIII-Tubulin expression. Mesoderm development was also assessed by enumeration of hematopoietic progenitors in the EBs.
The strong upregulation of HNF3β in the early stage EBs generated in serum suggested that serum might be important for the establishment, but not the maturation of the endoderm lineage. To test this possibility, EBs were initiated for 2 days in the presence of serum and then switched to serum-free conditions (SR). As shown in
It was further determined if these endodermal cells developed from brachyury+ or brachyury− cells. To address this question, GFP (Bry)+ and GFP (Bry)− cells were isolated from day 2.5 EBs by cell sorting. These populations were allowed to reaggregate and cultured as clusters until day 6. On day 6, they were moved to the tissue culture grade dishes in medium with bFGF for 4 days (total of 10 days). Gene expression analysis indicated that cells, which express HNF3β, segregated to the GFP+ fraction (d2.5) (
To further evaluate the liver potential of the brachyury expressing cells, cell populations derived from both the bry+ and bry− fractions were analyzed for expression of genes representing early hepatocyte development such as a-fetoprotein (AFP), albumin (ALB) and transthyretin (TTR) and genes indicative of maturation of the lineage including alphal-antitrypsin (AAT), tyrosine aminotransferase (TAT) and carbamoyl phosphate synthetase I (CPase). β-actin expression was used as a control. Cells were analyzed prior to sorting and subsequent to sorting into bry+ and bry− populations. Fetal liver and adult liver controls were also analyzed. Expression of all of these genes was restricted to the cells derived from the bry+ population, indicating that cells with hepatocyte characteristics develop from brachyury expressing cells.
The preliminary kinetic analysis described in Example 6 demonstrated that subpopulations of mesoderm with distinct developmental fates were generated in a defined temporal fashion. A more detailed kinetic analysis showed the dynamic development of the GFP+Flk1− (hereafter referred to as the GFP+ population) population between days 2.5 and 4.0 of differentiation (
To further investigate the kinetics of endoderm development, GFP+ cells from day 3.0 and 4.0 EBs were isolated and cultured under conditions that promote the differentiation into hepatocyte-like cells. As shown in
To further evaluate the endoderm potential of the GFP+ population, GFP+ and GFP− day 2.5 EB cells were cultured for 14 day days under conditions known to promote hepatocyte differentiation and then transplanted under the kidney capsule of recipient SCID-beige mice. Several mice were sacrificed immediately following the transplant, the kidney with the graft was sectioned and the sections were stained with antibodies against Hep 1 and AFP. Hep1 is a specific marker of hepatocytes whereas AFP is expressed in definitive endoderm and immature cells of the hepatocyte lineage. Some of the cells within the section stained positive for Hep1, whereas other cells, in the vicinity of the Hep 1+ cells were found to express AFP. No Hep1+ or AFP+ cells were found in the graft of the GFP− negative cells. These findings support PCR data and demonstrate that there culture conditions support the development of cells with characteristics of immature hepatocytes, as defined by express of Hep1.
While these transplantation experiments demonstrate the presence of hepatocyte-like cells in the grafts immediately following transplantation, it was difficult to monitor the maturation of these populations over time, as the transplanted tissues generated tumor-like masses known as teratomas. The teratomas likely develop from contaminating undifferentiated ES cells or from GFP+ primordial germ cells that are known to be of mesoderm origin and to express brachyury.
To further enrich for cells with endodermal potential, the effects of growth factors known to induce this cell population in other model systems were tested. Studies in Xenopus have shown that activin will induce both mesoderm and endoderm from ectoderm in culture. Of particular interest was the observation that activin behaved as a morphogen in this model in that it induced different cell types at different concentrations used. To determine if activin displayed similar potential in the ES/EB system, it was added to the EB cultures using the following protocol. ES cells were differentiated for 2 days in Stem Pro 34 medium without serum. At this stage, the developing EBs were harvested and recultured in IMDM supplemented with serum replacement (serum free) and activin at a concentration of 100 ng/ml. EBs were harvested at different days and assayed for GFP+ expression and expression of genes indicative of ectoderm, mesoderm and endoderm development. As shown in FIGS. 22A,B this amount of activin-induced brachyury as measured by GFP. While the kinetics of GFP induction was delayed compared to EBs differentiated in serum, this concentration of activin did induce substantial numbers of brachyury positive cells (60%) by day 6 of differentiation. Molecular analysis indicated that the activin-induced cells expressed a broad spectrum of genes associated with endoderm development including HNF3β, Mixl-1, Sox17, Hex-1, and pdx-1 (
To determine if activin displayed morphogenic properties in the ES differentiation model, different concentrations of the factor were added to the EB cultures. As little as 1 ng/ml of activin induced GFP+ expression (10% of the total population) by 7 days of culture (
Activin-induced EBs did not express genes associated with hematopoietic commitment indicating that this developmental program was not induced. To further evaluate the hematopoietic potential of these EBs, they were analyzed for progenitor potential. As expected from the molecular analysis, these EBs contained no appreciable numbers of primitive (Ep) or definitive (mac/mix) hematopoietic progenitors (
PCR analysis demonstrated that activin induced expression HNF3β as well as other genes known to be involved in endoderm differentiation. To better estimate the proportion of endodermal progenitors in the activin-induced EBs, cells from cultures stimulated with 100 ng/ml activin were stained with antibodies against HNF3β and Hex1. EBs from un-induced cultures were used as controls (Activin-). A significant portion of the activin induced population (estimated at 50-60% of total) expressed both HNF3β and Hex1. None of the cells in the un-induced EBs expressed these proteins. These findings clearly demonstrate that a substantial number of cells within these EBs are of the endoderm lineage.
To further investigate the potential of these activin-induced populations, GFP+ cells were isolated from EBs stimulated with 3 ng or 100 ng and cultured further (14 day total) in hepatocyte conductions. As shown in
If the teratomas generated by the serum-induced brachyury+ cells resulted from the presence of primordial germ cells, it is possible that the activin induced cells may represent a better source of progenitors for transplantation, as the germ-cell program may not be induced under these conditions. To test this hypothesis, GFP+ and GFP− cells were isolated from EBs induced with 100 ng/ml of activin and cultured for 14 days to promote the differentiation of hepatocyte-like cells. Following this culture period, the cells were harvested and transplanted to the kidney capsule of recipient animals. Three weeks following transplantation, the mice were sacrificed and the kidneys analyzed. Results are shown in FIGS. 26A-C. All mice engrafted with GFP− cells developed large, multilineage teratomas, consisting of cells from all three germ layers. In contrast, no teratomas were detected in the animals transplanted with GFP+ cells. These cells give rise to differentiated cell masses consisting of endodermal and mesodermal derived tissues including gut epithelium, bone and skeletal muscle. In some instances, skin was also observed in the graft from the GFP+ cells, suggesting that this lineage might also develop from a bry+ cell. These findings indicate that it is possible to generate GFP+ populations that give rise to differentiated tissue without forming teratomas following transplantation.
The foregoing examples clearly indicate that the hepatocyte lineage develops from a bry+ cell population that has both mesoderm and endoderm potential. Analysis of the bry+ fraction revealed that a subpopulation of these cells expressed the receptor tyrosine kinase c-kit (
The PCR analysis presented in
As a further assessment of the potential of activin treated cells, day 6 EBs differentiated in the presence of different concentrations of this factor were transferred to serum replacement media for 4 days and then replated in hepatocyte conditions for an additional 4 days. At day 14 of culture, the cells from each group were harvested and subjected to PCR expression analysis. Expression of Myf5 and skeletal actin were monitored to evaluate skeletal muscle development representing an additional mesoderm-derived lineage. Surfactant protein C (SP-C), a lung-specific gene, was included as a marker of endoderm differentiation in addition to AFP and ALB. As shown in
Brachyury positive and negative populations isolated from EBs stimulated with low and high concentrations of factor were also analyzed for expression of the skeletal muscle and endoderm genes. As shown in
In order to track and isolate endodermal as well as mesodermal populations during EB differentiation, ES cells which already have GFP targeted to the brachyury locus were targeted with human CD4 (hCD4) protein into the HNF3 β locus. HNF3 β is known to be expressed in most endodermal cell types, starting very early during embryonic development. Kaestner et al. (1994) Genomics 20:177-85. The targeting construct contains the human CD4 cDNA, truncated at amino acid number 424, and lacks most of the intracellular domain. This cell surface molecule was chosen because it cannot transduce signals and there are readily available antibodies for flow cytometry, which recognize hCD4 and will not cross react with endogenous mouse CD4. The vector also contains two regions of homology to the mouse HNF3 β locus, the bovine growth hormone poly(A) sequence, and a loxp flanked puromycin cassette as depicted in
The short arm and long arm of homology were obtained from a published HNF3B targeting vector. Weinstein et al. (1994) Cell 78: 575-588. The same long arm was used while a 1.8 kb fragment of the short arm was used to create the hCD4 knock in vector. The short arm contains part of the HNF3β promoter, exon 1, intron 1, and part of exon 2, ending just before the endogenous ATG initiation site. The long arm consists of a 7 kb sequence starting 3 kb downstream of the end of the last exon. Using standard molecular biology techniques, hCD4 with a kozak sequence followed by the bovine growth hormone poly(A) sequence along with a puromycin cassette was cloned in between the two arms of homology as shown in
The GFP-Bry ES cells described in example 2 were electroporated with the HNF3 β targeting vector. The cells were then selected with puromycin and clones were picked and expanded. Genomic DNA was isolated, cut with the restriction enzyme Sac I, and a Southern blot analysis was performed with a probe upstream of the endogenous ATG as indicated in
HNF3β is expressed transiently in EBs during serum induced differentiation. To determine whether hCD4 expression in CD4-HNF/GFP-T ES cells accurately reflects expression of the HNF3β gene during EB development, CD4 versus GFP expression was assessed.
A typical expression pattern of GFP versus hCD4 during a 5 day serum induced EB differentiation is shown in
During embryogenesis, both the mesoderm and endoderm lineages are derived from a structure called the primitive streak. Beddington and Robertson (1999) Cell 96: 195-209. The primitive streak is characterized by expression of specific genes such as Brachyury, which is expressed along the whole streak. HNF3β is also expressed in the streak, though it has only been detected in the anterior section. HNF3β expression as demonstrated by hCD4 was seen to be heterogeneous in the Bry-GFP+ population. It is possible that serum induction allows development of a group of primitive streak like cells, with the GFP+ HNF3β10 cells representing the posterior streak while the GFP+ HNF3βhi cells correspond to the anterior streak.
To determine if GFP and hCD4 expression could be used to separate posterior and anterior primitive streak like cells from CD4-HNF/GFP-T ES cells, EBs differentiated for 3.25 days were sorted into three populations: GFP+ hCD4lo, GFP+ hCD4med, and GFP+ hCD4hi (
Lineage tracing experiments have demonstrated that the anterior and posterior primitive streak develop into distinct lineages. The posterior streak gives rise to the yolk sac and hematopoietic cells while the anterior streak forms endoderm and anterior mesoderm such as cardiac tissue. Robb and Tam (2004) Sem. Cell & Dev. Biol. 15: 543-554. To determine if the primitive streak like cells from EBs behave functionally like the endogenous primitive streak, CD4-HNF/GFP-T ES cells were differentiated for 3.25 days, sorted for GFP+ hCD4lo and GFP+ hCD4hi populations and assayed for hematopoietic, cardiac, and endoderm potential. For all assays, the sorted populations were allowed to reaggregate for 48 hours in serum free conditions.
Studies in mice have demonstrated the importance of Wnt signaling as well as nodal/activin signaling in gastrulation and endoderm/mesoderm induction. Liu et. al. (1999) Nat. Genet. 22: 361-4. Example 12 hereinafter demonstrates that activin can induce both mesodermal and endodermal populations from ES cells in serum free cultures. To determine the effects of activin and Wnt signaling on the differentiation of ES derived primitive streak like cells, CD4-HNF/GFP-T cells were differentiated in serum free culture with either recombinant Wnt3a or activin A. Both activin and Wnt can initially induce GFP+ CD4+ cells, though Wnt displays a faster kinetics (
Considering that Wnt stimulation induces nodal and activin treatment induces Wnt3 (
Example 12 demonstrated that high levels of activin could induce endodermal gene expression from EBs differentiated in completely serum free conditions. Nodal, a key molecule involved in mesoderm/endoderm induction in vivo, signals through the same receptors as activin. It is also known that Nodal as well as inhibitors for Wnt signaling are expressed preferentially in or around the anterior primitive streak. Considering that the endoderm forms from the anterior primitive streak, it is possible that nodal/activin signaling in combination with inhibition of Wnt signals induces endoderm formation. To test this hypothesis, CD4-HNF/GFP-T ES cells were differentiated for 3.25 days in serum and GFP+ hCD4med cells were sorted and reaggregated in serum free media with no factor, DKK1 (Wnt inhibitor), activin, or activin and DKK1. The effect of these culture conditions on hCD4-HNF3β and GFP-Bry levels were analyzed by flow cytometry (