WO2018183653A1

WO2018183653A1 - Method of generating and using cd34+ cells derived from fibroblasts

Info

Publication number: WO2018183653A1
Application number: PCT/US2018/025106
Authority: WO
Inventors: Jalees Rehman; Asrar Malik
Original assignee: The Board Of Trustees Of The University Of Illinois
Priority date: 2017-03-29
Filing date: 2018-03-29
Publication date: 2018-10-04

Abstract

Methods of reprogramming mammalian fibroblasts to hemangioblast-like CD34+ mesodermal progenitor cells and further generating endothelial cells and erythroblasts by modulating Sox17 or Sox18 expression are provided as are methods of using the cells to form blood vessels and erythroblasts.

Description

METHOD OF GENERATING AND USING CD34⁺ CELLS

DERIVED FROM FIBROBLASTS

Introduction

[0001] This patent application claims benefit of priority from U.S. patent application number 62/478,387, filed March 29, 2017, the content of which is hereby incorporated by reference in its entirety.

[0002] This invention was made with government support under grant number HL118068 awarded by the National Institutes of Health. The government has certain rights in the invention.

Background

[0003] Terminally differentiated somatic cells have been used to generate induced pluripotent stem cells (iPSCs) via overexpression of a defined set of transcription factors. The most widely used set of reprogramming factors, Oct4 (Octamer-binding transcription factor 4), Sox2 (SRY-Box 2), Klf4 (Kruppel Like Factor 4) and c-Myc (MYC Proto- Oncogene) , collectively referred to as "OSKM, " was identified initially by screening 24 pre-selected factors in mouse embryonic fibroblasts (MEFs; Takahashi & Yamanaka (2006) Cell 126 ( 4 ) : 663-76 ) . Exogenous expression of OSKM has been shown to work for different types of somatic cells and for different species, including rhesus monkey (Liu, et al. (2008) Cell Stem Cell 3(6):587-90) and human cells (Takahashi, et al. (2007) Cell 131 (5) : 861-72) . Moreover, selected MEF-derived iPSCs expressing OSKM and having reduced expression of stem cell antigen 1 (Seal) and cluster of differentiation gene 34 (CD34) are suggested to exhibit higher reprogramming efficiency due to their hyper- energetic state (US 2018/0044642) . However, intermediate generation of iPSCs, which require further differentiation, is time consuming and has the added risk of teratoma formation caused by any residual pluripotent cells.

[0004] Direct conversion of a somatic cell such as a fibroblast into other cell types for personalized organ and tissue regeneration overcomes the drawbacks associated with iPSCs. Studies have shown that fibroblasts can be directly reprogrammed into hepatocyte-like cells, cardiomyocytes , neurons, endothelial cells (ECs) and erythroid progenitors. See, e.g., Capellera-Garcia, et al. (2016) Cell Rep. 15 ( 11 ): 2550-62. One approach relies on de-differentiation in which Oct-4, KLF4, Sox2, and c-Myc are expressed to partially de-differentiate fibroblasts, followed by applying endothelial growth factor VEGF to induce terminal endothelial differentiation, thereby avoiding formation of iPSCs (Kurian, et al. (2013) Nat. Methods 10:77-83; Li, et al. (2013) Arterioscler. Thromb . Vase. Biol. 33:1366-75; Margariti, et al . (2012) Proc. Natl. Acad. Sci . USA 109:139793-8; US 2014/0162366). The second approach relies on targeted over-expression of embryonic transcription factors known to regulate formation of endothelial cells. However, neither approach has been shown to be coupled to the generation of erythroblasts , which are critical for oxygen transport into ischemic tissue.

Summary of the Invention

[0005] This invention provides a method for producing endothelial cells by contacting a population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells with Soxl7 or Soxl8 and inducing endothelial cell differentiation, e.g., by culturing the population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells for about 10 days in endothelial cell growth medium supplemented with Bone Morphogenetic Protein 4 and Vascular Endothelial Growth Factor to produce endothelial cells. In some embodiments, the mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells are produced by contacting mammalian fibroblasts with 0ct4, Klf4, Sox2 and c-Myc; culturing the mammalian fibroblasts under conditions to generate a population of hemangioblast-like CD34⁺ cells; and isolating the population of hemangioblast-like CD34⁺ cells from the mammalian fibroblasts.

[0006] This invention also provides a method of treating ischemia or regenerating tissue, which involves the steps of contacting a population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells with Soxl7 or Soxl8; and administering to a subject in need of treatment an effective amount of the population of mammalian fibroblast- derived, hemangioblast-like CD34⁺ cells. In some embodiments, the population of mammalian fibroblast- derived, hemangioblast-like CD34⁺ cells is produced by contacting mammalian fibroblasts, e.g., autologous mammalian fibroblasts, with 0ct4, Klf4, Sox2 and c-Myc; culturing the mammalian fibroblasts under conditions to generate a population of hemangioblast-like CD34⁺ cells; and isolating the population of hemangioblast-like CD34⁺ cells from the mammalian fibroblasts.

[0007] This invention further provides a method for producing erythroblasts , which involves the steps of contacting a population of mammalian fibroblasts with an inhibitor of Soxl7 or Soxl8 expression; reprogramming the mammalian fibroblasts of (a) to produce a population of hemangioblast-like CD34⁺ cells; isolating the population of hemangioblast-like CD34⁺ cells from the mammalian fibroblasts; and inducing erythroblast differentiation, e.g., by culturing the population of fibroblast-derived, hemangioblast-like CD34⁺ cells for about 4 days in basal medium supplemented with Stem Cell Factor, Interleukin-3 , Erythropoietin, Insulin-like Growth Factor 1 and FMS-like tyrosine kinase 3 ligand. In some embodiments, the step of reprogramming the mammalian fibroblasts is achieved by contacting the mammalian fibroblasts with Oct4, Klf4, Sox2 and c-Myc; and culturing the mammalian fibroblasts under conditions to generate a population of hemangioblast-like CD34⁺ cells.

[0008] As a further aspect, this invention provides a method of treating an anemia by contacting a population of mammalian fibroblasts, e.g., autologous mammalian fibroblasts, with an inhibitor of Soxl7 or Soxl8 expression; reprogramming the mammalian fibroblasts to produce a population of hemangioblast-like CD34⁺ cells; isolating the population of hemangioblast-like CD34⁺ cells from the mammalian fibroblasts; and administering to a subject with an anemia an effective amount of the population of fibroblast-derived, hemangioblast-like CD34⁺ cells thereby treating the subject's anemia. In some embodiments, the step of reprogramming the mammalian fibroblasts involves contacting the mammalian fibroblasts with 0ct4, Klf4, Sox2 and c-Myc; and culturing the mammalian fibroblasts under conditions to generate a population of hemangioblast-like CD34⁺ cells.

Brief Description of the Drawings

[0009] FIG. 1 provides a schematic of method for converting human dermal fibroblasts (Fib) into induced endothelial cells and induced erythroblasts.

[0010] FIG. 2 shows the permeability of iECs in response to thrombin. Trans-endothelial electrical resistance (TER) was measured for 90 minutes in fibroblasts and iECs. Thrombin (3.5 pg/ml;0.5 unit) was added and the time course of resistance in fibroblasts and iECs was determined and normalized to the baseline. Statistical analysis of percentage of deduction of resistance in corresponding with thrombin in fibroblasts and iECs is shown. N=4.

[0011] FIG. 3 shows telomerase activity of iECs. Telomerase activity was measured by a telomeric repeat amplification protocol (TRAP) assay. TRAP analysis of positive control, negative control, fibroblasts (Fib) , ECs and iECs was carried out. Statistical analysis of telomerase activity in these cell types is shown. N=3. Data are presented as mean ± SE; **P<0.01.

[0012] FIG. 4 shows quantitative PCR analysis of hemoglobin beta and hemoglobin gamma by Taqman® assay in the erythroid colonies. N=3. Data are presented as mean ± SE. ***P<0.001, compared to fibroblasts, using Student's t test with multiple comparisons (Bonferroni Correction) .

[0013] FIG. 5 shows quantitative PCR analysis of transcriptional factors of hematopoietic cells - GATA1, RUNX1 and Pu.l in the fibroblasts (Fib) and induced erythroblasts (iErythroblasts) . N=3. Data are presented as mean ± SE; *P<0.05, **P<0.01, compared to fibroblasts, using Student's t test with multiple comparisons

(Bonferroni Correction) .

[0014] FIG. 6 shows that overexpression of Soxl7 and Soxl8 increased endothelial induction, as represented by VE- cadherin, CD31 and Flkl expression, when compared to control GFP overexpression. Data are presented as mean ± SE. *P<0.05 compared to control (scramble shRNA or GFP); ^# P<0.05 compared to cells overexpressing Soxl8; ^SP<0.05 compared to cells overexpressing Soxl7. Detailed Description of the Invention

[0015] Methods for generating multipotent mesodermal progenitors from de-differentiated adult fibroblasts have now been developed to provide a means of forming new blood vessels along with hematopoietic cells including erythroblasts . The methods involve reprogramming of human fibroblasts into endothelial cells and erythroblasts via the generation of intermediate hemangioblast-like cells. The cells generated in this manner have the potential to oxygenate ischemic tissue. In this approach, fibroblasts were first partially de-differentiated to generate the multipotent CD34⁺ hemangioblast-like cells (as opposed to pluripotent stem cells) via expression of OKSM factors and then subsequently converted into functional endothelial cells, iECs, and erythroblasts/erythroblasts by modulating the expression of Soxl7 and/or Soxl8. The generated iECs expressed the prototypic endothelial cell surface proteins, proliferated and formed tube-like structures, and responded normally to inflammatory stimuli by increasing expression of cell surface adhesion molecules. Fibroblast-derived human iECs also crucially restored endothelial barrier function underscoring their reparative potential. Implantation of human hemangioblast-like cells in NOD/SCID mice gave rise to human endothelial cells capable of forming blood vessels carrying erythroblasts, and they markedly improved human microvessel generation and cardiac function after myocardial infarction. Notably, it was shown that Soxl7 and Soxl8 were required for endothelial lineage conversion and that over-expression of Soxl7/Soxl8 markedly enhanced the generation of endothelial cells from fibroblasts, whereas suppression of Soxl7/Soxl8 facilitated the generation of erythroblasts. [0016] Accordingly, this invention provides methods for producing endothelial cells and/or erythroblasts by modulating the expression/level of Soxl7 and/or Soxl8 in mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells. SRY-related HMG-box (Sox) proteins are a family of DNA-binding transcription factors. Among Sox family members, Sox7, Soxl7, and Soxl8 are highly related and constitute the Sox subgroup F (SoxF) . Soxl7 and Soxl8 participate in various developmental processes and biologic activities, such as vascular development (Matsui, et al . (2006) J. Cell Sci. 119(Pt. 17 ): 3513-26) . Moreover, it has been shown that Soxl7 also plays an important role in fetal hematopoiesis in the yolk sac and fetal liver, especially in the maintenance of fetal and neonatal hematopoietic stem cells (HSCs), but not adult HSCs (Kim, et al. (2007) Cell 130(3) : 70-83) . Overexpression of Soxl7 has also been shown to confer fetal HSC characteristics onto adult hematopoietic progenitors (He, et al . (2011) Genes Dev. 25 ( 15 ): 1613-27 ) . Sox7 and Soxl8 are transiently expressed in hemangioblasts and hematopoietic precursors, respectively, at the onset of blood specification. Sustained expression of Sox7 and Soxl8, but not Soxl7, in early hematopoietic precursors from mouse embryonic stem cells and embryos enhances their proliferation while blocking their maturation (Gandillet, et al . (2009) Blood 114 (23) : 4813-22; Serrano, et al . (2010) Blood 115 ( 19 ): 3895- 98) .

[0017] Modulation of Soxl7 and/or Soxl8 expression includes increasing the expression/level of Soxl7 and/or Soxl8 and decreasing the expression/level of Soxl7 and/or Soxl8 protein. Increases in Soxl7/Soxl8 expression/levels can be achieved by transfecting a cell with a gene encoding Soxl7/Soxl8 or transducing a cell with Soxl7/18 protein using nucleic acid and protein sequences well known in the art. For example, nucleic acid molecules encoding human Soxl7 and Soxl8 are readily available under GENBANK Accession Nos . NM_022454 and NM_018419, respectively. Likewise, protein sequences for human Soxl7 and Soxl8 are known in the art and available under GENBANK Accession Nos. NP_071899 and NP_060889, respectively.

[0018] Plasmids, inducible vectors, lentiviral vectors, adenoviral vectors, and AAV vectors designed to express the Soxl7 and Soxl8 proteins in mammalian cells are readily available from commercial sources such as GenScript, Vector BioLabs, and Addgene . Expression Soxl7 and/or Soxl8 can occur transiently (e.g., using a non-integrative or episomal nucleic acid) or stably in a cell. Episomal or non-integrative nucleic acid molecules are not part of the chromosomal DNA and replicate independently thereof. Thus, during "transient expression" the transfected gene is not transferred to the daughter cell during cell division. Since its expression is restricted to the transfected cell, expression of the gene is lost over time. In contrast, stable expression of a transfected gene can occur when the gene is co-transfected with another gene that confers a selection advantage to the transfected cell. Such a selection advantage may be a resistance towards a certain toxin that is presented to the cell.

[0019] Transient increases in Soxl7 and/or Soxl8 levels can also be achieved by protein transduction. "Protein transduction" refers to the internalization from the external environment of proteins into a cell. As such, when a cell is transduced with a protein, the protein transits the cell membrane so as to pass from the external environment of the cell into the cell, e.g., into the cytoplasm of the cell. Protein transduction can be achieved by fusing a protein of interest, e.g., Soxl7 or Soxl8, to a protein transduction domain (PTD), which facilitates transport of the protein of interest into the cell. PTDs may include short cationic peptides (e.g., 5 to 100 amino acids, or from 5 to 25 amino acids in length) that can bind to the cell surface through electrostatic attachment to the cell membrane and be taken up by the cell by membrane translocation (Kabouridis (2003) Trends Biotech.

21 ( 11) : 498-503) . A given transduction protein molecule may include a single PTD or multiple PTDs (e.g., dimer, trimer, tetramer, pentamer, hexamer, heptamer, octamer, nonamer, decamer or larger multimer) or different PTDs can be conjugated to the protein of interest.

[ 0020 ] Specific PTDs include, but are not limited to, PTDs derived from human immunodeficiency virus 1 (HIV-I) TAT (Ruben, et al . (1989) J. Virol. 63:1-8); the herpes virus tegument protein VP22 (Elliott & O'Hare (1997) Cell 88:223- 233) ; the homeotic protein of Drosophila melanogaster Antennapedia (Antp) protein (Penetratin PTD; Derossi, et al. (1996) J. Biol. Chem. 271:18188-18193); the protegrin 1 (PG-I) antimicrobial peptide SynB (Kokryakov, et al . (1993) FEBS Lett. 327:231-236); and the Kaposi fibroblast growth factor (Lin, et al . (1995; J. Biol. Chem. 270:14255-14258) . Synthetic PTDs can also be used including, but not limited to, transportan (Pooga, et al . (1988) FASEB J. 12:67-77), MAP (Oehlke, et al. (1998) Biochim. Biophys. Acta. 1414:127-139), KALA ( yman, et al. (1997) Biochemistry 36:3008-3017) and other cationic peptides, such as, for example, various β-cationic peptides (Akkarawongsa, et al . (2008) Antimicrob. Agents Chemother. 52 ( 6) : 2120-2129) and poly-histidine, poly-lysine or poly-arginine peptides. Additional PTD peptides and variant PTDs also are provided in, for example, US 2005/0260756, US 2006/0178297, US 2006/0100134, US 2006/0222657, US 2007/0161595, US 2007/0129305, EP 1867661, WO 2000/062067, WO 2003/035892, WO 2007/097561 and WO 2007/053512.

[0021] Transduction proteins can be fabricated using any convenient protocol. The PTD can be conjugated to a protein of interest using any convenient protocol, such as, for example, conjugation by recombinant means or by chemical coupling. The linkage of the components in the conjugate can be by any convenient method, so long as the attachment of the linker moiety to the protein of interest does not substantially impede the desired activity of the protein of interest. Linkers and linkages that are suitable for chemically linked conjugates include, but are not limited to, disulfide bonds, thioether bonds, hindered disulfide bonds, and covalent bonds between free reactive groups, such as amine and thiol groups .

[0022] Decreases in Soxl7/Soxl8 expression/levels can be achieved using an inhibitor of Soxl7 or Soxl8 expression. An "inhibitor of Soxl7 or Soxl8 expression" refers to a molecule the measurably decreases the mRNA or protein levels of Soxl7 or Soxl8 by at least 30%, 40%, 50% 60%, 70%, 80%, 90% or 100% as compared to a cell not contacted with said inhibitor. Suitable inhibitors of Soxl7 or Soxl8 expression include inhibitory RNAs such as antisense, ribozymes, short interfering nucleic acid (siNA) , short interfering RNA (siRNA) , double-stranded RNA (dsRNA) , micro-RNA (miRNA) , and short hairpin RNA (shRNA) molecules capable of mediating RNA interference (RNAi) against Soxl7 or Soxl8 nucleic acid sequences. For example, miR-141 has been shown to downregulate the expression of Soxl7 (Jia, et al. (2012) J. Mol. Diagn. 14 ( 6) : 577-85 ) , whereas miR-7a and miR-24-3p are suggested to suppress the expression of Soxl8 (Olbromski, et al . (2016) Oncol. Rep. 36 ( 5 ) : 2884-2892 ) . Further, siRNA targeting the sequences 5'-

GAGCUAAGCAAGAUGCUAG (SEQ ID NO : 1 ) , 5' -GGCUGUUCAAAAAUUUCGG (SEQ ID NO: 2), 5 ' -GAACCCAGAUCUGCACAAC (SEQ ID NO : 3 ) of Soxl7 have been used to decrease expression of Soxl7 (Chew, et al. (2011) J. Neurosci. 31 (39) : 13921-35) , whereas siRNA targeting position 1344-1362 ( 5' -CUCUCUCAUACGCGUGUAU; SEQ ID NO: 4) of human SOX18 mRNA effectively inhibits Soxl8 expression (Wang, et al. (2015) Oncol. Rep. 34(3) : 1121-8). Further exemplary siRNA molecules are disclosed herein and include the sequences 5 ' -GGACCGCACGGAAUUUGAA (SEQ ID NO: 5), 5' -GCAUGACUCCGGUGUGAAU (SEQ ID NO: 6) and 5'-

CCGCGGUAUAUUACUGCAA (SEQ ID NO : 7 ) targeting Soxl7. In certain embodiments, an shRNA or siRNA is used to inhibit the expression of Soxl7 or Soxl8.

[0023] In one aspect of this invention, a method is provided for producing endothelial cells by contacting a population of mammalian fibroblast-derived, hemangioblast- like CD34⁺ cells with Soxl7 or Soxl8 and inducing endothelial cell differentiation.

[0024] In another aspect, this invention provides a method for producing erythroblasts by contacting a population of mammalian fibroblasts with an inhibitor of Soxl7 or Soxl8 expression; reprogramming said mammalian fibroblasts to produce a population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells; isolating the population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells from the mammalian fibroblasts; and inducing erythroblast differentiation .

[0025] For the purposes of this invention, the term "hemangioblast-like" refers to a cell that is phenotypically similar to a hemangioblast and is capable of differentiating into both a blood lineage cell, in particular a erythroblast, and a vascular endothelial cell. [0026] A "population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells" refers to a population of hemangioblast-like CD34⁺ cells derived from mammalian fibroblasts. The term "derived from," when referring to cells, indicates that the cells were obtained from the stated source at some point in time. In some cases, cells derived from a given source will undergo differentiation/de-differentiation such that the original cells no longer exist, but the continuing cells will be understood to be derived from the same source. In the present case, hemangioblast-like CD34⁺ cells are derived from mammalian fibroblasts by reprogramming the mammalian fibroblasts to produce a population of hemangioblast-like CD34⁺ cells. As used herein, the term "reprogramming" refers to the process of partially de-differentiating a terminally differentiated cell (i.e., a fibroblast) into a cell exhibiting multipotent characteristics, i.e., a cell that can develop into more than one cell type, but is more limited than pluripotent cells.

[0027] Fibroblasts refers to a type of cell in connective tissue that produces the extracellular matrix and collagen but do not produce bone, bone minerals, or cartilage. Fibroblasts are large, flat, elongated (spindle-shaped) cells and may be defined as having high expression of FSP- 1, collagen III, and/or collagen type 15. Fibroblasts of use in this invention can be readily isolated from the skin of a mammal, such as a human, non-human, primate, mouse, rat, rabbit, horse, pig, cow, dog, cat, or other companion animal. In certain embodiments the fibroblasts are adult human dermal fibroblasts.

[0028] Mammalian fibroblasts, which are typically devoid of CD34 expression, can be reprogrammed or partially dedifferentiated to CD34⁺ cells using any suitable method described in the art. In particular embodiments, mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells are obtained by contacting mammalian fibroblasts with Oct4, Sox2, Klf4 and c-Myc (i.e., the OSKM quartet); culturing the mammalian fibroblasts under conditions to generate a population of hemangioblast-like CD34⁺ cells; and isolating the population of hemangioblast-like CD34⁺ cells from the mammalian fibroblasts. As with Soxl7 and Soxl8, the step of contacting mammalian fibroblasts with the OSKM quartet can carried out by conventional transfection or protein transduction approaches. For example, to expedite expression, polycistronic lentiviral vectors have been used to deliver the OSKM quartet to somatic cells in a single lentiviral construct reducing the number of genomic insertions (Kim, et al . (2009) Stem Cells 27(3):543-9; Chang, et al. (2009) Stem Cells 27 (5) : 1042-9) . Alternatively, the OSKM reprogramming factors have been delivered with minimal or total absence of genetic modifications by using LoxP sites and Cre-induced excision and piggyBac transposon excision of integrated reprogramming vector sequences (Soldner, et al. (2009) Stem Cells 136 (5) : 964-77; Kaji, et al . (2009) Nature 458 (7239) : 771-5) , or an oriP/EBNAl-based episomal vector (Yu, et al. (20090 Science 324 (5928 ): 797-801) . As another strategy, poly-arginine tagged Oct4, Sox2, Klf4 and c-Myc proteins have been found to readily enter cells and translocate into the nucleus (Zhou, et al . (2009) Cell Stem Cell 4(5): 381-4). Using a similar approach, protein-induced pluripotent stem cells have been obtained from human newborn fibroblasts after several rounds of treatment with cell extracts of HEK293 cell lines expressing poly-arginine tagged OSKM genes (Kim, et al. (2009) Cell Stem Cell 4(6):472-6). In some embodiments, the mammalian fibroblasts are further contacted with miR302-367 , Lin28 and/or an shRNA against p53 to facilitate CD34⁺ generation. See, e.g. , Kurian, et al . (2013) Nature Methods 10:77-83.

[0029] Upon contacting the mammalian fibroblasts with the OSKM quartet, the mammalian fibroblasts are cultured under conditions to generate a population of hemangioblast-like CD34⁺ cells. The terms "culture," "culturing," "grow," "growing," "maintain," "maintaining," etc., when referring to cell culture itself or the process of culturing, can be used interchangeably to mean that a cell is maintained outside the body (e.g., ex vivo) under conditions suitable for survival. Cultured cells are allowed to survive, and culturing can result in cell growth, differentiation, or division. The term does not imply that all cells in the culture survive or grow or divide, as some may naturally senesce, etc. Cells are typically cultured in media, which can be changed during the course of the culture. For the purposes of this invention, the mammalian fibroblasts are cultured at a temperature typically at about 37 °C with a C0₂ level of typically around 5% in an appropriate media (including salts, buffer, and nutrients) for a time sufficient for OSKM quartet to de-differentiate the fibroblasts to hemangioblast-like CD34⁺ cells. As described herein, a 2-day exposure to OKSM followed by 5 days in basal medium supplemented with bFGF (Basic Fibroblast Growth Factor; see, e.g., NCBI reference Gl : 153285461 ) was sufficient to de-differentiate fibroblasts to CD34⁺ progenitor cells. Suitable basal media for culturing the fibroblasts include, but are not limited to, DMEM, F12, or a combination thereof, supplemented with, e.g., serum or a serum replacement, L-glutamine, β-mercaptoethanol , and nonessential amino acids. [0030] Given that CD34 is a cell surface glycoprotein, the presence of hemangioblast-like CD34⁺ cells, i.e., cells with increased expression of CD34 compared to fibroblasts, can be readily isolated by conventional cell sorting techniques. In this respect, the population of hemangioblast-like CD34⁺ cells are isolated from the parental mammalian fibroblasts {i.e., CD34^~ cells). Cell sorting techniques of use include, for example, fluorescence-activated cell sorting (FACS) , magnet- activated cell sorting (MACS), and panning.

[0031] To induce endothelial cell differentiation, hemangioblast-like CD34⁺ cells are cultured in cell growth medium appropriate for endothelial cell growth {i.e., endothelial cell growth media) . Ideally, the endothelial cell growth medium includes endothelial cell growth factors such as BMP4 (Bone Morphogenetic Protein 4) and VEGF (Vascular Endothelial Growth Factor) . The course or duration of endothelial cell differentiation is typically in the range of 6 and 20 days, e.g., at least 6, 8, 10, 12 or 14 days and can optionally be carried out in the presence of collagen (e.g., the cells are grown on collagen plates) . In particular embodiments, the hemangioblast-like CD34⁺ cells are cultured for about 10 days in endothelial cell growth medium.

[0032] To induce erythroblast differentiation, hemangioblast-like CD34⁺ cells are cultured in cell growth medium appropriate for erythroblast growth. Ideally, the erythroblast growth medium includes a basal medium {e.g., IMDM and Ham's F12) supplemented with EPO, a key regulator in ex vivo erythropoiesis methodologies, along with other growth factors such as SCF (Stem Cell Factor) , IL-3

( Interleukin-3 ) , IGF-1 (Insulin-like Growth Factor 1) and Flt-3-L (FMS-like tyrosine kinase 3 ligand) . See Singh, et al. (2014) Adv. Regen . Med. Article ID 426520. The course or duration of erythroblast differentiation is typically in the range of 3 and 7 days, e.g., at least 3, or 4 days and can optionally be carried out in the presence of collagen (e.g., the cells are grown on collagen plates) . In particular embodiments, the hemangioblast-like CD34⁺ cells are cultured for about 4 days and erythroblasts are collected as a non-adherent cell fraction.

[0033] One of skill will also appreciate that the methods of this invention are typically carried out with a plurality of hemangioblast-like CD34⁺ cells, to form a population of cells that, over the course of differentiation, includes increasing numbers of endothelial cells or erythroblasts. Preferably, the percentage of endothelial cells or erythroblasts in the population of cells increases to about 50-100% over the course of differentiation, e.g., more than 65, 75, 80, 85, 90, 95 or higher percentage endothelial cells or erythroblasts. In some embodiments, the endothelial cells or erythroblasts can be further separated from the population using the methods described herein and prepared for storage (e.g., in freezing media) or therapeutic application. Endothelial cell surface markers that can be used for separation include VE-cadherin, endoglin (CD105), vWF, CD31, and Z01 (tight junction protein) , though negative selection can also be used (e.g., alpha-S A or other non-endothelial cell marker) . To confirm differentiation, endothelial cells can be assessed for expression of endothelial markers VE- cadherin, CD31, FLK1, and vWF. Similarly, erythroblasts can be assessed for expression of erythroid marker glycophorin CD235a, as well as RUNXl, PU.l, GATAl, human hemoglobin beta and hemoglobin gamma. [0034] Standard methods known in the art may be used to determine the detectable expression, low expression or lack thereof of the various markers discussed herein. Suitable methods include, but are not limited to, immunocytochemistry, immunoassays, flow cytometry, such as FACS, and polymerase chain reaction (PCR) , such as reverse transcription PCR (RT-PCR) . Suitable immunoassays include, but are not limited to, western blot analysis, enzyme- linked immunoassays (ELISA) , enzyme-linked immunosorbent spot assays (ELISPOT assays) , enzyme multiplied immunoassay techniques, radioallergosorbent (RAST) tests, radioimmunoassays, radiobinding assays and immunofluorescence. Western blot analysis, ELISAs and RT- PCR are all quantitative and can be used to measure the level of expression of the various markers if present. Antibodies and fluorescently-labelled antibodies for all of the various markers discussed herein are commercially- available .

[0035] This invention also provides methods for treating a disease or condition or regenerating tissue in a subject wherein generation of endothelial cells and/or erythroblasts would provide a benefit. Such diseases or conditions include, e.g., ischemia and anemia. Accordingly, one aspect of the invention provides a method for treating ischemia or regenerating tissue by contacting a population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells with Soxl7 or Soxl8; and administering to a subject in need thereof an effective amount of the population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells. Representative ischemic diseases or conditions include, but are not limited to, limb ischemia, congestive heart failure, cardiac ischemia, kidney ischemia and ESRD, stroke, and ischemia of the eye. Representative uses for tissue regeneration include bone regeneration, cardiac regeneration, vascular regeneration, and neural regeneration, as well as tissue engineering and tissue repair .

[0036] In another aspect, the invention provides a method for treating an anemia by contacting a population of mammalian fibroblasts with an inhibitor of Soxl7 or Soxl8 expression; reprogramming said mammalian fibroblasts to produce a population of hemangioblast-like CD34⁺ cells; isolating the population of hemangioblast-like CD34⁺ cells from the mammalian fibroblasts; and administering to a subject with an anemia an effective amount of the population of fibroblast-derived, hemangioblast-like CD34⁺ cells. Representative anemic conditions that can be treated in accordance with the invention include, but are not limited to, anemia associated with kidney disease, e.g., end-stage renal disease; anemia of chronic disease; anemia of cancer; chemotherapy-induced anemia; iron deficiency anemia; anemia associated with sickle cell disease; and the like.

[0037] In certain embodiments, the mammalian fibroblasts used in the methods of the invention are obtained from a donor source (allogenic) or are autologous mammalian fibroblasts, e.g., obtained via skin biopsy.

[0038] The cell-based compositions described herein can be administered as a pharmaceutically or physiologically acceptable preparation or composition containing a physiologically acceptable carrier, excipient, or diluent, and administered to the tissues of the recipient subject, including humans and non-human animals. Representative compositions can be prepared by resuspending the cells in a suitable liquid or solution such as sterile physiological saline or other physiologically acceptable injectable aqueous liquids. The amounts of the components to be used in such compositions can be routinely determined by those having skill in the art.

[0039] The cells or compositions thereof can be administered by placement of cell suspensions onto absorbent or adherent material, e.g., a collagen sponge matrix, and insertion of cell-containing material into or onto the site of interest. Alternatively, the cells can be administered by parenteral routes of injection, including subcutaneous, intravenous, intramuscular, and intrasternal . Other modes of administration include, but are not limited to, intranasal, intrathecal, intracutaneous, percutaneous, enteral, and sublingual. In one embodiment of the present invention, administration of the cells can be mediated by endoscopic surgery.

[0040] For injectable administration, the cells or compositions thereof is in sterile solution or suspension or can be resuspended in pharmaceutically- and physiologically-acceptable aqueous or oleaginous vehicles, which may contain preservatives, stabilizers, and material for rendering the solution or suspension isotonic with body fluids (i.e., blood) of the recipient subject. Non-limiting examples of excipients suitable for use include water, phosphate buffered saline, pH 7.4, 0.15 M aqueous sodium chloride solution, dextrose, glycerol, dilute ethanol, and the like, and mixtures thereof. Illustrative stabilizers are polyethylene glycol, proteins, saccharides, amino acids, inorganic acids, and organic acids, which may be used either on their own or as admixtures . The amounts or quantities, as well as the routes of administration used, are determined on an individual basis, and correspond to the amounts used in similar types of applications or indications known to those of skill in the art. [0041] The cells or compositions thereof can be administered to body tissues, including liver, pancreas, lung, salivary gland, blood vessel, bone, skin, cartilage, tendon, ligament, brain, hair, kidney, muscle, cardiac muscle, nerve, skeletal muscle, joints, and limbs.

[0042] The number of cells in a cell suspension or composition and the mode of administration may vary depending on the site and condition being treated. As non- limiting examples, in accordance with the present invention, about 3.5xl0⁷ to 3.0xl0⁸ cells are injected to effect tissue repair. Consistent with the Examples disclosed herein, a skilled practitioner can modulate the amounts and methods of cells-based treatments according to requirements, limitations, and/or optimizations determined for each case.

[0043] The following non-limiting examples are provided to further illustrate the present invention.

Example 1 : Experimental procedures

[0044] Cell Culture and Transfection . Normal human adult dermal fibroblast cells (Lonza) and human neonatal dermal fibroblast cells (Lonza) were cultured in Dulbecco ' s Modified Eagle's Medium (DMEM; Life Technologies) containing 10% fetal bovine serum (FBS), 50 ϋ/ml penicillin and 50 mg/ml streptomycin, 2.25 mM each L-glutamine and non-essential amino acids. Fibroblasts (5.0xl0⁵/100 mm dish) and Platinum-A (Plat-A) cells (Cell Biolabs Inc.)

(5. OxloVlOO mm dish) were plated in DMEM medium without antibiotics .

[0045] Dishes (100 mm) were plated with Plat-A cells, one dish per plasmid, each encoding 0CT4, KLF4, SOX2, and c-MYC (i.e., OKSM, all obtained from Addgene) , as well as green fluorescent protein (GFP) as a control. On Day 0, Fugene® HD transfection reagent (Roche) and 10 μg of the appropriate plasmid DNA were added to one 100 mm dish cultured with Plat-A cells. The virus-containing medium was collected and filtered from dishes of transfected Plat-A cells on Day 1 and Day 2 for titration.

[0046] The medium of human fibroblasts was replaced with virus-containing medium on Day 1 and Day 2. On Day 3 the medium was switched to DMEM/F12 (Life Technologies) , 20% KnockOut™ serum replacement (Life Technologies), 10 ng/ml basic fibroblast growth factor (bFGF; R&D Systems) , 1 rtiM GlutaMAX™ (L-alanyl-L-glutamine) , 0.1 mM non-essential amino acids, and 55 μΜ β-mercaptoethanol with medium changes every day for 5 days to continue the de- differentiation phase. The fibroblasts transduced by OKSM and maintained in this medium were referred to as "dedifferentiated fibroblasts" or "De-Diff-Fib . " The protocol is shown in FIG. 1. All cell lines were maintained in an incubator (37°C, 5% C0₂) .

[0047] Isolation of Hemangioblast-Like Cells from De^¬ differentiated Fibroblasts . Hemangioblast-like cells were identified by the presence of CD34, a marker of human progenitors, including endothelial progenitor cells and hematopoietic progenitors (Sidney, et al . (2014) Stem Cells 32:1380-1389). Hemangioblast-like CD34⁺ cells generated during de-differentiation of fibroblasts were isolated with anti-CD34-conjugated magnetic beads (Miltenyi) according to manufacturer's instructions with slight modifications. Briefly, up to 10⁹ cells were incubated at 4°C with 100 μΐ of Fc-blocking solution and 100 μΐ anti-CD34 magnetic beads in a total volume of 500 μΐ FACS blocking buffer. After 30 minutes, cells were sorted by two consecutive rounds of column separation through applying MACS® separation magnets to increase purity. [0048] Differentiation of Hemangioblast-Like Cells into Endothelial Cells (iECs) . The isolated hemangioblast-like cells were plated on 0.1% gelatin-coated plates

( 1.5xl0⁵/well in 6-well plate) and cultured in endothelial medium EGM-2. The medium was supplemented with 50 ng/ml bone morphogenetic protein 4 (BMP4; R&D Systems) and 50 ng/ml Vascular endothelial growth factor (VEGF; R&D Systems) (FIG. 1) to induce endothelial lineage specification and the medium was changed every other day. Upon reaching -90% confluence, cells were split 1:3. Primary human aortic endothelial cells or lung microvascular endothelial cells were purchased from Lonza and cultured in EGM-2 medium as human endothelial control cells. All cell lines were maintained in an incubator (37°C, 5% C0₂) .

[0049] Differentiation of Hemangioblast-Like Cells into Erythroblasts . To generate erythroid cells, sorted hemangioblast-like cells were plated in 0.1% gelatin-coated plates ( 1.0xl0⁶/well in a 6-well plate) and cultured for 4 days in erythroblast induction medium composed of 49%

(Iscove's Modified Dulbecco's Medium (IMDM; Life Technologies), 49% Ham's F12 (Life technologies), 1% Insulin-Transferrin-Selenium-Ethanolamine (ITS-X) , 1% lipid

(Life Technologies), 5 mg/100 ml I-ascorbic acid (Sigma), 500 mg/100 ml BSA (Sigma), and 200 μΜ 1-thioglycerol

(Sigma) , supplemented with 100 ng/ml Stem Cell Factor (SCF; PeproTech) , 10 ng/ml Interleukin-3 (IL-3; PeproTech) , 5 U/ml Erythropoietin (EPO; R&D Systems), 40 ng/ml Insulinlike Growth Factor 1 (IGF-1; Sigma) , 1 μΜ Dexamethsone

(Sigma) , 50 ng/ml FMS-like tyrosine kinase 3 ligand (Flt3- L; PeproTech) (FIG. 1) . As is known in the art, EPO is a key regulator in ex vivo erythropoiesis methodologies along with other growth factors such as SCF, IL-3, IGF-1, and Fit-3. See Singh, et al . (2014) Adv. Regen . Med. Article ID 426520. Unlike adherent endothelial cells, erythroblasts detached upon generation and were collected as non-adherent cell fraction at Day 4 for characterization.

[0050] Gene Expression and Bioinformatics. Gene expression data on fibroblasts, hemangioblast-like cells derived from fibroblast de-differentiation, induced endothelial cells

(iECs) and adult human endothelial cells (ECs) were obtained using the Affymetrix® GeneChip® Human Transcriptome Array 2.0. The analysis was performed in the R environment for statistical computing using the Affy and Limma packages. The 11 Affymetrix CEL files were pre- processed by Robust MultiChip Analysis. For each gene, the normalized intensity in each cell type was paired with the normalized intensity of the remaining samples and analyzed with the Limma package. A linear model was fitted to the design matrix for the samples using Limma, and differentially expressed genes identified for each pairwise comparison using a moderated t statistic, which applies an empirical Bayes method to borrow information across genes. P-values were adjusted for multiple testing using the Benjamini and Hochberg' s method, which controls the false discovery rate. Gene expression profiles were visually depicted using heatmaps . Samples and genes were ordered by hierarchical clustering and represented by the dendrogram. The color bar indicates gene expression levels in a binary logarithm (log2) scale. The green to red color spectrum represented low to high expression levels.

[0051] Matrigel™ Plug in vivo Angiogenesis Assay. To test the ability of iECs to form blood vessels, an in vivo Matrigel™ plug assay was performed. Immune-deficient NOD- SCID mice (NOD . Cq-Prkdcscid I12rgtmlWj 1/SzJ; Jackson Laboratory) were used to avoid the rejection of human iECs. Anesthesia was induced using a mixture of xylazine (Rompun® 2%, Bayer) at 10 mg/kg body weight and ketamine (Imalgene 1000®, Merial) at 100 mg/kg body weight in NaCl at 0.9% i.p. The animals' backs were shaved and swabbed with hexomedine . Prior to injection, hemangioblast-like cells were harvested using 0.05% trypsin/EDTA (Invitrogen) . A total of lxlO⁶ cells (either fibroblast-derived hemangioblast-like cells, control human fibroblasts, or control human primary ECs) were suspended in 250 μΐ of cold Matrigel™ (Matrigel™ basement-membrane matrix from BD adjusted to 9.8 mg/ml with phosphate buffered saline (PBS) ) . The cell-containing Matrigel™ solutions were then injected subcutaneously in the abdomen of mice, carefully positioning the needle between the epidermis and the muscle layer. Seven days later, mice were sacrificed and the Matrigel™ plugs were removed by a wide excision of the skin, including the connective tissues.

[0052] Myocardial Infarction Surgery and Intramyocardial Injection of Cells. Immune-deficient NOD-SCID mice were purchased from Jackson Laboratory (NOD . Cq-Prkdcscid I12rgtmlWj1/SzJ) . Anesthesia was introduced with 1.5-3% isoflurane inhalation in a closed glass chamber and Etomidate (10 mg/kg body weight, i.p.) . Mice were orally intubated with one 18G Angiocath™ sleeve and artificially ventilated with a rodent respirator (tidal volume 0.2-0.3 ml (based on body weight) , rate 135 strokes/minute) . Surgical anesthesia was maintained using 1% isoflurane delivered through a vaporizer with air connected in series to rodent ventilator. A dose of buprenorphine sustained release (0.1 mg/kg s.c.) was administered pre-operatively . Under microscopic view, left thoracotomy was performed by 1 cm careful incision along sternum and 1 mm to the left from a midline between the 2^nd and 4^th rib in layers (skin, pectoral muscles and ribs avoiding mammary arteries). Having the heart in view, the pericardium was removed to ligate the left main coronary artery with 8-0 Prolene® suture 1-2 mm below the ostium. Immediately after ligation, 50 μΐ of concentrated cells, i.e., lxlO⁶ fibroblast-derived, hemangioblast-like CD34⁺ cells or control fibroblasts suspended in Matrigel™ (Matrigel™ basement-membrane matrix from BD) , were injected into the border zone of the infarct at 2-3 different areas using a gastight 1710 syringe (Hamilton). The cells were pre-incubated with 2.5 pg/ml Dil stain (1,1* -Dioctadecyl-3 , 3 , 3 ' , 3 ' -

Tetramethylindocarbocyanine Perchlorate; Thermo Fisher Scientific) for 5 minutes. The chest cavity was then closed in three layers: intercostal muscles-interruptive, pectoral muscles and skin-continuous with the 7-0 nylon sutures. The thorax was drained with a PE-10 cannula inserted into incision upon closure to evacuate the chest cavity and reestablish proper intrapleural pressure. After tube removal, mice were weaned from anesthesia and placed in the warming pad until ambulatory and then returned to the appropriate housing facility. The mouse surgeon was blinded to the nature of the injected cells.

[ 0053 ] Transthoracic Echocardiography. Evaluation of regional cardiac function based on left ventricle (LV) thickness and morphology was performed by transthoracic echocardiography (ECG) . Mice were anesthetized by 3.5% isoflurane and placed a heated platform (VisualSonics , Inc.) where ECG (lead II), heart rate, respiratory waveform and respiratory rate were monitored and displayed on the ultrasound's CPU display. Then, three views were recorded with the sample volume in the proper modes (B-mode, M-mode, pulsed Doppler or tissue Doppler) : 1. The parasternal long axis [B-mode and/or M-mode]; 2. The parasternal short axis [Mmode was taken at the papillary level followed by tissue Doppler of the myocardium]; 3. Apical view [pulsed Doppler of the mitral flow] . Fractional shortening (FS) and ejection fraction (EF) were used as indices of cardiac contractile function (Ubil, et al. (2014) Nature 514:585- 590) . M-mode tracings were used to measure LV internal diameter at end diastole (LVIDd) and end systole (LVIDs) . FS was calculated according to the following formula: FS (%) = [(LVIDd - LVIDs) /LVIDd] ^χ 100. EF is estimated from (LVEDV-LVESV) /LVEDV 100%, wherein LVESV is Left ventricular end systolic volume and LVEDV is left ventricular end diastolic volume.

[0054] Quantitative Real-Time PCR. Total RNA was extracted using the PureLink® RNA Mini Kit (Ambion) according to manufacturer's protocol. cDNA was synthesized by High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) . cDNA (50 ng relative to RNA amount) was amplified by real-time PCR. FastStart™ Universal SYBR® Green Mastermix (Roche) was used for quantitative real-time PCR. The PCR primers used in this analysis are listed in Table 1.

TABLE 1

Endothelial

R 5' -CATTGAACAACCGATGCGTGA 19 Cadherin

Cluster of F 5' -AACAGTGTTGACATGAAGAGCC 20

Differentiation CD31

R 5' -TGTAAAACAGCACGTCATCCTT 21

31

Fetal Liver F 5' -TCGAAGCATCAGCATAAGAAACTT 22

Flkl

Kinase 1 R 5' -GCCCACTGGATGCTGCA 23

Von Willebrand F 5' -GGGCTGCCAGAAACGCT 24 v F

Factor R 5' -CAAGATACACGGAGAGGCTCACT 25

F 5' -TCTGCCTCCTCCACGAAG 26

SRY-Box 17 Soxl7

R 5' -ACGCCGAGTTGAGCAAGA 27

F 5' -AGCTCCTTCCACGCTTTG 28

SRY-Box 18 Soxl8

R 5' -GTGTGGGCAAAGGACGAG 29

GATA Binding F 5 ' -ATCAAGCCCAAGCGAAGACT 30

GATA1

Protein 1 R 5' -CATGGTCAGTGGCCTGTTAAC 31

SPI1 Proto- F 5' -GGGGTGGAAGTCCCAGTAAT 32

Pu.1

Oncogene R 5' -ACGGATCTATACCAACGCCA 33

Runt Related F 5' -AGAACCTCGAAGACATCGGC 34 Transcription RUNX1

R 5' -GGCTGAGGGTTAAAGGCAGTG 35 Factor 1

[0055] TaqMan® Gene Expression Assays from Applied Biosystems were used to detect human hemoglobin beta and human hemoglobin gamma with the primer/probe sets of hemoglobin beta (Hs00747223_gl) and hemoglobin gamma

(Hs00361131_gl) , respectively, while Glyceraldehyde 3- phosphate dehydrogenase (GAPDH, Hs02758991_gl) served as control .

[0056] Cytospin Preparation and Hema 3™ Staining. Cells (10^s) were suspended in 300 ml and used for analysis with the Thermo Scientific Shandon™ 4 Cytospin® kit. The slides were fixed and stained with Hema 3™ kits (Fisher Scientific) according to manufacturer' s protocol to visualize erythroblasts . The cells were imaged using a Zeiss ApoTome inverted microscope with 20X, 40X and 60X obj ectives .

[0057] Immunocytochemistry, Immunohistochemistry and

Fluorescence Microscopy. Immunocytochemistry was performed on cells after fixation with 4% paraformaldehyde. Fixed cells were permeabilized and blocked with 5% normal goat serum, 0.1% Triton™ X-100 (nonionic surfactant) in PBS for 30 minutes at room temperature. Then the cells were incubated with primary antibodies for 2 hours and secondary antibodies labeled with fluorescent Alexa Fluor™ 488 and Alexa Fluor™ 647 dyes (Invitrogen, Carlsbad, CA) for 1 hour. Immunohistochemistry was performed on paraffin- embedded sections after heat-induced antigen retrieval (30 minutes at 95°C in 0.01 M sodium citrate, pH 6) . The slides were permeabilized and blocked with 5% normal goat serum, 0.1% Triton™ X-100 in PBS for 1 hour at room temperature. The slides were then incubated with primary antibody at 4°C overnight. For immunofluorescence, fluorescent Alexa Fluor™ 488 dye and Alexa Fluor™ 568 dye (Invitrogen, Carlsbad, CA) labeled secondary antibodies were applied to identify the primary antibodies. The slides were mounted in ProLong® Gold antifade reagent with DAPI (Invitrogen) . Florescence images were acquired using Zeiss ApoTome microscope with 20X, 40X and 60X objectives. Confocal image acquisition was performed using a Zeiss LSM 780 laser scanning microscope (Carl Zeiss Jena) with 20X, 40X and 60X water objectives. Primary antibodies used were mouse anti-human CD31 (DAKO, Denmark) , goat anti-VE-cadherin (Santa Cruz Biotech) , rat anti-mouse TER 119/Erythroid cells (BD Pharmingen) , mouse anti-human CD235a (Glycophorin A; R&D Systems) , mouse anti- human CD233 (Thermo Fisher Scientific) .

[ 0058 ] Flow Cytometry. Cells were harvested from culture dishes and analyzed on the CyAn™ ADP instrument (Beckman Coulter) with Summit software 4.4. The antibodies used for flow cytometry analysis were CD34-PE (Miltenyi Biotec) , Alexa Fluor™ 647 mouse anti-human CD144 (BD Biosciences) , PE anti-hCD31/PECAM-l (R&D Systems), Alexa Fluor™ 647 mouse anti-human CD309 (VEGFR-2; BD Biosciences), PE mouse anti- human CD106 (BD Biosciences) , and APC mouse anti-human CD62E (BD Biosciences). Isotype controls were Alexa Fluor™ 647 mouse IgGlk (BD Biosciences), Alexa Fluor™ 647 mouse IgGlk (BD Biosciences), PE mouse IgGlk (BD Pharmigen) , PE mouse IgG2bk (BD Biosciences), FITC mouse IgGlk (BD Pharmigen) and APC mouse IgG2ak (BD Biosciences). Cell proliferation was assayed by flow cytometry using the APC BrdU flow cytometry kit (BD Biosciences) according to manufacturer's instructions.

[0059] Western Blot Analysis. Immunoblotting was performed as previously described (Zhang, et al. (2013) Cell 153:216- 227) . The primary antibodies applied in western blot analysis were mouse anti-actin (Sigma); goat anti-VE- cadherin (Santa Cruz); rabbit anti-Flkl (VEGFR2; Cell Signaling Technology) , mouse anti-human CD31 (DAKO, Denmark) , goat antihuman Soxl7 (R&D Systems) , and goat anti-human Soxl8 (R&D Systems) . Donkey horseradish peroxidase (HRP) -linked anti-goat, mouse and rabbit IgG (Santa Cruz) were used as secondary antibodies. Signals were developed by enhanced chemiluminescence SuperSignal™ West Femto Maximum Sensitivity Substrate (Thermo Scientific) and detected with the LAS-3000 mini imaging system (Fuji Film, Japan) .

[0060] Transendothellal Electrical Resistance Measurement . Assessment of responses of iEC to inflammatory stimuli was performed by measuring thrombin-induced increase in endothelial permeability according to known methods (Tiruppathi, et al . (2000) Proc. Natl. Acad. Sci. USA 97:7440-7445). Human iECs and control human ECs (5. OxIO cm²) were cultured on gelatin-coated gold electrodes for 1 day. The smaller electrode and larger counter electrode were connected to a phase-sensitive lock- in amplifier to monitor the voltage. A constant current of 1 μΑ was supplied by a 1 V, 4000 Hz AC signal connected serially to a 1 ΜΩ resistor between the smaller electrode and the larger counter electrode. Before each experiment, confluent endothelial monolayers using both cell types were incubated in serum-free medium for 1 hour. Thrombin (3.5 pg/ml, 0.5 Units) was added as indicated. The data are presented as a change in the resistive portion of electrical impedance normalized to its value at time 0.

[0061] Telomerase Activity. Telomerase activity was determined using the telomeric repeat amplification protocol (TRAP) . TRAP assay was performed using the TRAPeze® Telomerase Detection kit (Millipore) and a Cy5 fluorescently labeled TS primer according to established protocols (Herbert, et al . (2006) Nat. Protoc. 1:1583- 1590). Densitometry of telomerase-specific ladder and internal standard were quantified using ImageJ and used to calculate relative telomerase activity compared to an internal standard control.

[0062] Tube Formation Assay. A cell suspension containing 0.3xl0⁵/well of either ECs (control) or iECs was placed on top of wells coated with Matrigel™ (10 mg/ml; BD Matrigel™ Basement Membrane Matrix) in a 24-well plate (Nunc) . Rearrangement of cells and formation of capillary structures were observed at 24 hours.

[0063] Plasmid Constructs _r Retrovirus or Lentivirus Packaging, and Infection. Small double-stranded hairpin siRNAs targeting Soxl7 were designed by BLOCK-iT™ RNAi Designer ( Invitrogen) , synthesized by Integrated DNA Technologies (IDT), and inserted into a pLL3.7 lentiviral vector. The target sequences of Soxl7 siRNAs were Soxl7-l: 5' -GGACCGCACGGAATTTGAA (SEQ ID NO: 36); Soxl7-2: 5'- GCATGACTCCGGTGTGAAT (SEQ ID NO:37); Soxl7-3: 5'- CCGCGGTATATTACTGCAA (SEQ ID NO:38). Lipofectamine™ 2000 was used to transfect pLL3.7 lentiviral vector to HEK293T cells (seed 2.0xl0⁶/6-well dish 1 day prior) . The medium containing lentiviruses was collected and filtered two days after transfection. The iECs was incubated with the virus cocktail supplemented with 8 μg/ml polybrene for 6-8 hours. Soxl8 siRNA lentivectors (i023437a, i023437b, i023437c) and scrambled siRNA GFP lentivectors (LV015-G) were purchased from Applied Biological Materials Inc. Packaging of Soxl8 siRNA lentivectors or scrambled siRNA GFP lentivectors was performed according to Applied Biological Materials' instructions. To overexpress Soxl7/Soxl8 and distinguish overexpressed genes from endogenous gene expression, full- length Soxl7 (pMXs-Soxl7) and Soxl8 (pMXs-Soxl8) plasmids were purchased from Addgene . To package retrovirus expressing Soxl7 and Soxl8, Fugene® HD transfection reagent (Roche) was used to transfect Plat-A cells ( 1.0xl0⁶/6-well dish) . The medium containing the retrovirus was collected and filtered two days after transfection. For Soxl7 and Soxl8 over-expression studies, iECs were incubated with a virus cocktail supplemented with 4 μg/ml polybrene for 6-8 hours .

[ 0064 ] Colony-Forming Assay. MethCult™ H4034 Optimum with Recombinant Cytokines and EPO were purchased from Stemcell Technologies. A colony forming assay was performed according to the manufacturer instructions to evaluate the colony-forming function of generated erythroblasts . Briefly, 3xl0³ induced erythroblasts or fibroblasts were collected, washed and then mixed in 3x1300 μΐι MethCult™. The suspended cells were dispensed into petri dishes using a syringe and blunt-end needle. The dishes were then incubated for 14 days in a humidified incubator at 37 °C and 5% C0₂. The erythroid colonies were evaluated and counted using an inverted microscope and gridded scoring dishes. The colonies were plucked for further assays.

[0065] Statistics. Values were reported as mean ± standard error. Significant differences between means were determined by paired t-test when two groups were compared or one-way ANOVA using Bonferroni corrections for multiple comparisons with the GraphPad Prism 5 software. Statistical significance was set at p < 0.05.

Example 2 : De-differentiation of Human Fibroblasts into CD34⁺ Progenitor Cells

[0066] OCT4, KLF4, SOX2, c-MYC (OKSM) were expressed in human fibroblasts for 2 days as described herein (FIG. 1) . On Day 3, the medium was replaced with bFGF and other constituents for 5 days to generate CD34⁺ progenitor cells. The fibroblast-derived cells generated by this protocol were referred to as "de-differentiated fibroblasts" or "De- Diff-Fib" or "hemangioblast-like cells". Endothelial or erythroid growth factors were then applied to differentiate the De-Diff-Fib cells into endothelial cells or erythroblasts (FIG. 1) .

[0067] Over the 7-day de-differentiation period, the expression of CD34 gradually increased for about 2 to 47

(relative gene expression to B2M, normalized to Day 0) . OCT4 mRNA levels peaked at Day 3 to Day 4 (to about 180 relative to B2M, normalized to Day 0) and then steadily decreased (to approximately 55), indicative of circumventing pluripotency . Expression of SOX2, c-MYC and KLF4 exhibited similar expression changes as OCT4. Expression of NANOG, which is an endogenous pluripotency factor that was not part of the OKSM cocktail, was also assessed. NANOG mRNA was down-regulated during fibroblast de-differentiation (from about 1 to 0.15 relative to B2M, normalized to Day 0), showing that short-term 2-day expression of OKSM did not induce pluripotency . Upon assessing generation of hemangioblast-like CD34⁺ cells by flow cytometry, it was found that de-differentiation of both neonatal and adult human dermal fibroblasts gave rise to the CD34⁺ cells; with higher CD34⁺ yield when neonatal fibroblasts were de-differentiated as compared to adult fibroblasts (16.311.6 vs. 8.7+1.2). For subsequent studies, human adult dermal fibroblasts were analyzed because of their potential utility for vascular regeneration.

Example 3: Lineage Conversion of Fibroblast-Derived, Hemangioblast-Like CD34⁺ Progenitors into Endothelial Cells

[0068] CD34⁺ cells were isolated from the de-differentiated fibroblast population using magnetic bead sorting, and the cells were exposed first to endothelial lineage induction medium composed of standard endothelial growth medium supplemented with endothelial differentiation factors VEGF and BMP-4 for 10 days. During the endothelial induction phase, gene expression of endothelial cell markers VE- cadherin, vWF, CD31 and Flkl progressively increased by about 15- to 40-fold. Flow cytometry of cells at Day 10 of induction demonstrated that 8% of cells were positive for either VE-cadherin or CD31 and 5% were positive for both surface markers. Western blot analysis established marked upregulation of proteins VE-cadherin, CD31, and Flkl in these generated cells termed induced endothelial cells (iECs) . Immunofluorescence staining demonstrated that the bulk of VE-cadherin was however localized in the cytoplasm as opposed to the cell surface membrane where VE-cadherin is typically found in confluent adherent endothelial cells (ECs) . The relatively high cytosolic VE-cadherin expression also explained why only 8% of iECs were positive for VE- cadherin by flow cytometry as flow cytometry only monitors the cell surface expressed protein. Since VE-cadherin trans-interactions between cells stabilizes VE-cadherin at cell membrane and promotes formation of adherens junctions (Daneshjou, et al . (2015) J. Cell Biol. 208:23-32), iECs were co-cultured with primary adult human ECs . Confocal imaging demonstrated that this co-culturing increased membrane localization of VE-cadherin in iECs such that iECs formed characteristic adherens junctions via homotypic interactions .

[0069] To identify gene expression signatures of de- differentiation and endothelial lineage conversion, transcriptomic analysis of adult fibroblasts, fibroblast- derived CD34⁺ cells, iECs, and primary human ECs was performed. A heat map of global gene expression revealed that iECs clustered with ECs, whereas CD34⁺ cells clustered with fibroblasts; thus, iECs exhibited a gene expression profile similar to adult primary ECs whereas fibroblast- derived CD34⁺ cells retained fibroblast signatures. Heat maps of EC-specific and fibroblasts-specific genes indicated that gain of endothelial identify and loss of fibroblast identity occurred during the transition from the CD34⁺ progenitor state to iECs.

Example 4 : Functional Responses of Generated iECs

[0070] Key functions of ECs are to form of a restrictive and plastic barrier and upregulate cell surface adhesion molecules in response to inflammatory stimuli. Thus, trans- endothelial electrical resistance (TER) was assessed in a confluent monolayer of iECs as a measure of barrier function and plasticity. In response to thrombin, a significant decrease in TER of iEC monolayers was observed whereas parent fibroblasts showed no response (FIG. 2) . Importantly, iECs also exhibited a recovery of TER, the restoration of barrier, typical of normal endothelium. It was also demonstrated that iECs formed capillary networks in 2D-Matrigel™. To assess responses to the proinflammatory cytokine TNF-a, expression of adhesion molecules CD106 (VCAMl) and CD62E (E-selectin) was measured. iECs responded with upregulation of VCAMl and E- selectin proteins and mRNA in the same manner as primary human ECs .

Example 5 : Proliferation Potential and Telomerase Expression in iECs

[0071] As the ability of ECs to proliferate in required for vascular regeneration, proliferation of iECs was assessed. This analysis demonstrated that 20-30% of iECs were labeled with BrdU, a value similar to proliferation of primary human ECs. As cell expansion induces cell senescence and results in cell cycle arrest due to continuous shortening of telomeres (Armanios (2013) J. Clin. Invest. 123:996- 1002; Blasco (2007) Nat. Chem. Biol. 3:640-649; Rando & Chang (2012) Cell 148:46-57), it was determined whether the generated iECs represented an aging cell population. Telomerase activity was determined since it maintains telomere length and protects against premature senescence in embryonic and neonatal cells (Armanios (2013) J. Clin. Invest. 123:996-1002; Blasco (2007) Nat. Chem. Biol. 3:640- 649; Rando & Chang (2012) Cell 148:46-57). It was observed that telomerase activity of iECs, as assessed by telomeric repeat amplification protocol (TRAP) , was significantly greater as compared in adult fibroblasts or primary adult human ECs (FIG. 3) . Importantly, despite higher telomerase activity, iECs did not show significantly elevated expression of the pluripotency transcription factor 0CT4 or the proto-oncogene c- YC.

Example 6: Fibroblast-Derived Generation of Erythroblasts

[0072] Fibroblast-derived, hemangioblast-like CD34⁺ progenitors were cultured in erythroblast induction medium for 4 days to assess generation of erythroblasts. Unlike iECs, which remained adherent to the substrate, emerging erythroblasts detached from the CD34⁺ cell layer. Hema 3™ staining Cytospin® preparations of induced erythroblasts showed varied morphology reflecting different stages of maturation from basophilic erythroblasts, polychromatic erythroblasts to mature erythroblasts. To address whether induced erythroblasts expressed erythroid markers, the Cytospin® preparations were stained for CD235a (GPA, Glycophorin A) and CD233 (band 3 of SLC4A1, solute carrier family 4 member 1) . The results indicated that induced erythroblasts, which were nucleated, expressed the late stage erythroblast surface membrane proteins CD235a and CD233. A few enucleated cells were also CD233⁺, indicating transition to mature erythroblasts. Flow cytometry quantification demonstrated that 4.2+0.8% of the cells undergoing erythroblast induction were CD235a⁺. Erythroid colony-forming-cell (CFC) assay demonstrated formation of distinct erythroid colonies (about 35 colonies/10000 cells vs. 0 colonies for fibroblasts) . Moreover, gene expression analysis showed expression of human hemoglobin beta and hemoglobin gamma confirming the erythroid nature of the generated human cells (FIG. 4) . Gene expression of hematopoietic transcriptional factors, RUNXl and Pu.l, as well as erythroblast transcription factor GATA1 (Hewitt, et al. (2014) Curr. Opin. Hematol. 21:155-164; Takeuchi, et al. (2015) Proc. Natl. Acad. Sci. USA 112:13922-13927) were also markedly enhanced in comparison to fibroblasts (FIG. 5) . These data indicate that fibroblast-derived CD34⁺ cells giving rise to functional ECs and erythroblasts are hemangioblast-like progenitor cells.

Example 7: SOX17 SOX18 Function as Switch Mechanism for Generation of iECs

[0073] The expression of vascular developmental genes was assessed in fibroblasts, hemangioblast-like cells, iECs, and primary adult ECs. Only Soxl7 and Soxl8 showed a pattern suggestive of their mechanistic role in endothelial fate transition. It was observed that the transcriptional regulator Soxl8 (a member of Sry-transcription factor family) was significantly up-regulated in hemangioblast- like cells, and even more so in iECs but was minimally expressed in the parent fibroblasts. Soxl7, another member of the Sry transcription factor family, showed similar upregulation in iECs. The transcription factor ETV2 (also known as ER71), which when over-expressed induces fibroblast-to-endothelial transition (Han, et al . (2014) Circulation 130:1168-1178; Morita, et al . (2015) Proc . Natl. Acad. Sci. USA 112:160-165), was not significantly up-regulated during lineage conversion of hemangioblast- like cells to iECs.

[0074] It was next investigated whether the transcription factors Soxl7 and Soxl8 were required for conversion of hemangioblast-like cells to iECs. Soxl7 and Soxl8 expression gradually increased during the 10 Day induction period. Immunoblot analysis confirmed upregulation of Soxl7 and Soxl8 in iECs. To assess whether expression of Soxl7 or Soxl8 was required for the induction of iECs, lentiviral shRNAs were used to deplete Soxl7 or Soxl8 during the endothelial induction phase. Using multiple shRNA constructs, a Soxl7 shRNA and Soxl8 shRNA were identified that achieved significant Soxl7 depletion and Soxl8 depletion, respectively. Depletion of Soxl7 prevented differentiation of hemangioblast-like cells to iECs as demonstrated by suppression of endothelial genes (VE- cadherin, CD31 and Flkl), both at mRNA and protein levels when compared to control shRNA. Depletion of Soxl8 also inhibited upregulation of VE-cadherin and CD31 but did not significantly affect Flkl levels. These data also showed that depleting either Soxl7 or Soxl8 using highly specific shRNAs inhibited the induction of Soxl7 or Soxl8, respectively, suggesting that Soxl7 and Soxl8 cross- activated each other during iEC differentiation from hemangioblast-like cells.

[ 0075 ] Based on these findings, it was subsequently determined whether over-expression of Soxl7 or Soxl8 would enhance endothelial lineage conversion. Hemangioblast-like cells undergoing endothelial induction were transfected with retroviruses encoding Soxl7 or Soxl8 on Day 1. Significant increases in mRNA and protein levels of VE- cadherin, CD31 and Flkl were observed when compared to the control group (FIG. 6). Quantitative analysis with flow cytometry showed that overexpression of either Soxl7 or Soxl8 also enhanced endothelial differentiation nearly 10- fold as evidenced by increased fraction of cells positive for cell surface VE-cadherin expression when compared to only using growth factors for endothelial lineage conversion (64.9+7.2% vs. 6.1+1.3% and 54.7+2.8% vs. 5.7+0.8%, respectively). Thus, conversion of hemangioblast- like cells into iECs is mediated by activation of transcription factors Soxl7 and Soxl8 and combining fibroblast de-differentiation with targeted over-expression of these two factors markedly augmented the efficacy of endothelial lineage conversion.

Example 8: SOX17/SOX18 Inhibit Generation of Erythroblasts from Hemangioblast-Like Cells

[0076] It was next examined whether Soxl7 and Soxl8 also function as key switches suppressing the transition of hemangioblast-like cells to erythroblasts. This analysis indicated that the expression of Soxl7/Soxl8 increased in hemangioblast-like cells when compared to the parent fibroblasts but progressively decreased during the induction of erythroblasts, consistent with the role of Soxl7/Soxl8 in favoring endothelial fate conversion at the hemangioblast-like cell bifurcation. To explore the regulatory role of Soxl7 and Soxl8 in the induction of erythroblasts, Soxl7 and Soxl8 were depleted by shRNA, respectively, in fibroblasts before de-differentiation. Depletion of Soxl7 or Soxl8 increased the expression of hematopoietic transcription factors RUNXl and Pu .1 as well as erythropoiesis-specific regulator GATA1 during erythroblast induction. Flow cytometry also confirmed increased erythropoiesis upon Soxl7 depletion trending towards increased erythropoiesis. In addition, it was observed that depleting Soxl7 or Soxl8 significantly increased the formation of colonies using the in vitro erythroid CFC assay (about 40-50 colonies/10000 cells for Soxl7 or Soxl8 shRNA vs. 30-35 colonies for scrambled shRNA) consistent with the essential role of Soxl7/Soxl8 in directing the fate of hemangioblast-like cells to either endothelial cells or erythroblasts that is, presence of Soxl7/Soxl8 favored endothelial lineage differentiation whereas their absence enhanced erythroid differentiation. Example 9: Hemangioblast-Like Cells Form Blood Vessels and Erythroblasts in vivo

[0077] The in vivo relevance of human fibroblast-derived hemangioblast-like cells was assessed. Hemangioblast-like cells were suspended in Matrigel™ and the suspensions were subcutaneously injected into immune deficient NOD-SCID mice. Matrigel™ plugs explanted after 1 week showed that implantation of either human ECs or hemangioblast-like cell-derived iECs induced the formation of blood vessels whereas blood vessels did not form following implantation of fibroblasts alone.

[0078] To evaluate whether the ECs lining newly formed blood vessels and erythroblasts localized in vessels of the Matrigel™ plug were derived from implanted human hemangioblast-like cells or host mouse, a human-specific anti-CD31 antibody was employed to stain the vascular endothelium and human specific anti-CD235a antibody was used to stain human erythroblasts, after confirming the species-specificity of the anti-CD235a antibody. Staining of explanted hemangioblast-like cells in Matrigel™ plugs demonstrated that the vessels stained positive for human CD31 and erythroblasts within these vessels stained positive for human CD235a. It was observed that the newly formed human blood vessels also contained both human erythroblasts and murine host erythroblasts, indicating communication between the mouse vasculature and newly formed vessels derived from iECs. In notable contrast, Matrigel™ plugs containing primary human ECs showed presence of erythroblasts derived from the host mouse and cells were negative for human CD235a, indicating that transplanted adult human ECs were incapable of giving rise to erythroblasts . [ 0079 ] A mouse myocardial infarction (MI) model was used to assess the regenerative potential of adult human hemangioblast-like cells in which Matrigel™ suspension containing either hemangioblast-like cells or control fibroblasts were injected in the peri-infarct area immediately distal to coronary artery ligation. Masson trichrome staining of cross section of hearts on Day 14 post-MI showed reduced fibrosis in hearts injected with hemangioblast-like cells. Heart weight/body weight ratio was also decreased in mice receiving hemangioblast-like cells, indicating suppression of adverse myocardial remodeling after infarction. Echocardiography demonstrated significantly improved cardiac contractility, as measured by ejection fraction and fractional shortening at 1 and 2 weeks post-MI in hearts receiving hemangioblast-like cells when compared to hearts which received control fibroblasts (Table 2). Hemangioblast-like cell implantation also improved morphologic and hemodynamic indexes as assessed by echocardiography (Table 2) .

TABLE 2

Fraction

HLC 75.23±6.37 55.48+14.06*** 43.50+18.70** (%)

Fractional Fib 42.37+5.29 16.07+4.25 11.34+3.89 Shortening

HLC 43.34+5.69 29.21+8.87*** 25.66+12.44*** (%)

Cardiac Fib 15.12+1.01 12.21+6.36 11.45+5.12 Output

HLC 15.33+1.20 18.20+3.69** 16.80+5.87* (ml/min)

LV Mass Fib 99.91+15.24 155.20+44.34 159.30+45.67 (mg) HLC 94.29+16.19 112.00+30.70* 136.8+51.47

LV Mass Fib 79.93+12.20 125.82+35.87 127.40+36.54 Corrected

HLC 75.43+12.96 89.73+24.53* 109.5+41.18 (mg)

LVIDs, LV internal diameter at end of systole; LVIDd, LV internal diameter at end of distole; LVESV, LV end systolic volume; LVEDV, LV end diastolic volume.

Data are presented as mean+SE. ***P<0.001, **P<0.01, *P<0.05, compared to fibroblasts treatment.

Fib, fibroblast; HLC, hemangioblast-like cells.

[0080] Immunofluorescence staining demonstrated that implantation of hemangioblast-like cells resulted in formation of human CD31⁺ blood vessels carrying human CD235a⁺ erythroblasts , indicating that implanted cells contributed to vascular regeneration and human erythropoiesis . Gene expression analysis of infarcted tissue implanted with CD34⁺ cells demonstrated expression of human CD31 and human VE-cadherin in contrast to the area implanted with control human fibroblasts.

[0081] The present studies establish a novel role for the transcription factors Soxl7/Soxl8 as critical endogenous regulators of the transition from fibroblasts to endothelial cells and erythroblasts via the de- differentiation to hemangioblast-like cells. Hemangioblast- like cells in situ gave rise to functional endothelial cells and erythroblasts and displayed profound regenerative effects when transplanted into the myocardium after infarction where they formed functional human blood vessels as well as erythroblasts. These cells may thus be of significant therapeutic value in ischemic disease with compromised tissue oxygenation because their progeny generated both the vasculature and enhanced oxygen-carrying capacity .

Claims

What is claimed is :

1. A method for producing endothelial cells comprising contacting a population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells with Soxl7 or Soxl8 and inducing endothelial cell differentiation thereby producing endothelial cells.

2. The method of claim 1, wherein the mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells are produced by

(a) contacting mammalian fibroblasts with Oct4, Klf4, Sox2 and c-Myc;

(b) culturing the mammalian fibroblasts under conditions to generate a population of hemangioblast-like CD34⁺ cells; and

(c) isolating the population of hemangioblast-like CD34⁺ cells from the mammalian fibroblasts.

3. The method of claim 1, wherein the step of inducing endothelial cell differentiation comprises culturing the population of mammalian fibroblast-derived, hemangioblast- like CD34⁺ cells for about 10 days in endothelial cell growth medium supplemented with Bone Morphogenetic Protein 4 and Vascular Endothelial Growth Factor to produce endothelial cells.

4. A method of treating ischemia or regenerating tissue comprising

(a) contacting a population of mammalian fibroblast- derived, hemangioblast-like CD34⁺ cells with Soxl7 or Soxl8; and (b) administering to a subject in need of treatment an effective amount of the population of mammalian fibroblast- derived, hemangioblast-like CD34⁺ cells thereby treating the subject's ischemia or regenerating tissue in the subject.

5. The method of claim 4, wherein the population of mammalian fibroblast-derived, hemangioblast-like CD34⁺ cells is produced by

(i) contacting mammalian fibroblasts with Oct4, Klf4, Sox2 and c-Myc;

(ii) culturing the mammalian fibroblasts under conditions to generate a population of hemangioblast-like CD34⁺ cells; and

(iii) isolating the population of hemangioblast-like CD34⁺ cells from the mammalian fibroblasts.

6. The method of claim 5, wherein the mammalian fibroblasts are autologous mammalian fibroblasts.

7. A method for producing erythrocytes comprising

(a) contacting a population of mammalian fibroblasts with an inhibitor of Soxl7 or Soxl8 expression;

(b) reprogramming the mammalian fibroblasts of (a) to produce a population of hemangioblast-like CD34⁺ cells;

(c) isolating the population of hemangioblast-like CD34⁺ cells from the mammalian fibroblasts; and

(d) inducing erythrocyte differentiation thereby producing erythrocytes .

8. The method of claim 7, wherein the step of reprogramming the mammalian fibroblasts comprises

(i) contacting the mammalian fibroblasts with Oct4, lf4, Sox2 and c-Myc; and (ii) culturing the mammalian fibroblasts under conditions to generate a population of hemangioblast-like CD34⁺ cells.

9. The method of claim 7, wherein the step of inducing erythrocyte differentiation comprises culturing the population of fibroblast-derived, hemangioblast-like CD34⁺ cells for about 4 days in basal medium supplemented with Stem Cell Factor, Interleukin-3 , Erythropoietin, Insulinlike Growth Factor 1 and FMS-like tyrosine kinase 3 ligand to produce erythrocytes.

10. A method of treating an anemia comprising

(b) administering to a subject with an anemia an effective amount of the population of fibroblast-derived, hemangioblast-like CD34⁺ cells thereby treating the subject's anemia.

11. The method of claim 10, wherein the step of reprogramming the mammalian fibroblasts comprises

(i) contacting the mammalian fibroblasts with Oct4, Klf4, Sox2 and c-Myc; and

(ii) culturing the mammalian fibroblasts under conditions to generate a population of hemangioblast-like CD34⁺ cells.

12. The method of claim 10, wherein the mammalian fibroblasts are autologous mammalian fibroblasts.