WO2018204762A1 - Methods and compositions of modeling diseases with hematopoietic cell systems - Google Patents

Abstract

The present invention relates to methods and compositions for developing a hematopoietic cell system to model diseases. These models are used to determine the efficacy of compounds to treat the disease.

Description

METHODS AND COMPOSITIONS OF MODELING DISEASES WITH

HEMATOPOIETIC CELL SYSTEMS CROSS REFERENCE TO RELATED APPLICATIONS

[0001] The presents application claims priority to U.S. provisional application 62/501,804, filed May 5, 2017, entitled METHODS AND COMPOSITIONS OF MODELING DISEASES WITH HEMATOPOIETIC CELL SYSTEMS, the contents of which are hereby incorporated by reference herein in its entirety.

SEQUENCE LISTING

[0002] The present application is being filed along with a Sequence Listing in electronic format. The Sequence Listing file, entitled SL_20931003PCT.txt, was created on May 3, 2018, and is 8,650 bytes in size. The information in electronic format of the Sequence Listing is incorporated herein by reference in its entirety.

FIELD OF THE INVENTION

[0003] The invention relates to methods and compositions for modeling diseases with hematopoietic cell systems. These models are used to determine the efficacy of compounds to treat the disease through a genomic signaling center.

BACKGROUND OF THE INVENTION

[0004] Blood is one of the few tissues in the human body that can be readily obtained from patients for analysis of primary diseased cells. Those having ordinary skill in the art have struggled to produce a sufficient number of CD34+ cells using normal expansion techniques for next generation sequencing techniques, such as chromatin immunoprecipitation sequencing (ChlP-seq), high-throughput chromatin immunoprecipitation (HiChIP), and chromatin interaction analysis by paired-end tag sequencing (ChlA-PET) from CD34+ progenitor cells. In previous studies, overexpression of c-MYC and BCL-XL in normal multipotent hematopoietic progenitor cells resulted in exponential replication of glycophorin A+ erythroblasts (GYPA+). See, Hirose et al, Stem Cell Reports, Vol. 1, 499-508 (2013), which is hereby incorporated by reference in its entirety. When an inducible expression system was used to control expression of these genes, removal of the inducible factor resulted in maturation of the cells through normal erythropoiesis. See, Hirose et al.

[0005] Blood disorders that inhibit normal erythropoiesis also prevent production of a sufficient number of cells for next generation sequencing techniques. Further, such disorders do not allow for analysis of different treatments for the disorder in a primary cell culture because cells usually die before completion of an experiment. Failure of erythropoiesis is a phenotype generally caused by a ribosomal gene haploinsufficiency. See, Narla et al, Blood, 115(16):3196- 3205, (2010); Lipton et al., Current Opinion Pediatrics, 22(1): 12-19, (2010); Ganapathi et al, British Journal of Hematology, 141(3): 376-387, (2008); Liu et al, Blood, 107(12):4583-4588, (2006), which are hereby incorporated by reference in its entirety. For example, Diamond- Blackfan anemia (DBA), is a disorder of the bone marrow in which the bone marrow is unable to produce a sufficient number of red blood cells. The genetic basis of the disorder is not completely clear; however, about one-quarter of DBA patients have a variety of mutations in RPS 19, a ribosomal protein gene. See Gazda et al, British Journal ofHaematology, 127(1): 105- 13 (2014), which is hereby incorporated by reference in its entirety.

[0006] Additional studies have sought to determine further information on the genetic basis of DBA. Western blot analysis of RPS19 protein expression in bone marrow CD34+ cells revealed three and two-fold decreases of the protein in two samples from patient disease cells to confirm that at the protein level that haplo-insufficiency is a cause of DBA in patients with one functional allele. On the other hand, compared to CD34+ cells, expression of the RPS19 protein in peripheral blood mononuclear cells did not differ from that of the control. Decreased expression of RPS 14 or RPS19 was also observed to cause p53 activation in human

hematopoietic progenitor cells. See, Dutt et al, Blood, 3;117(9):2567-76 (2011), which is hereby incorporated by reference in its entirety. p53 activation in animal models of ribosomal dysfunction has also been previously demonstrated. See Danilova et al, Blood, 112(13):5228- 5237 (2008); McGowan et al, Nature Genetics, 40(8):963-970 (2008); Barlow et al, Nature Medicine, 16(l):59-66 (2010); Sieff et al., British Journal of Haematology, 148(4):611-622 (2010), which are hereby incorporated by reference in their entireties.

[0007] Using gene set enrichment analysis, an experimentally determined set of p53 genes were demonstrated to be coordinately induced by RPS14 shRNAs compared with control shRNAs in genome-wide gene expression profiling experiments of primary cells. See, Dutt et al. Knockdown of either RPS14 or RPS19 induced the mRNA expression of well-established p53 target genes: p21 and BAX, and reduced the mRNA expression of GATA1 as shown by quantitative RT-PCR (qRT-PCR). See, Dutt et al., Blood, 3;117(9):2567-76 (2011), which is hereby incorporated by reference in its entirety. Activation of p53 was demonstrated to be sufficient to impair erythropoiesis and inactivation of p53 via a chemical compound was observed to rescue the effects of RPS14 or RPS19 deficiency in human hematopoietic progenitor cells. The treatment of hematopoietic progenitor cells with pifithrin-a (PFT-a), a compound that blocks the transcriptional transactivation activity of p53, also blocked the induction of p21 gene expression in response to knockdown of RPS 14 or RPS19.

[0008] Therefore, there is a long-felt need in the art to develop a disease model, e.g. a model of DBA or 5q- syndrome, that can conditionally rescue or induce the phenotype of the disease and that allows for production of a sufficient number of cells for next generation sequencing techniques to determine the effects of treatments on genomic signaling centers that control expression of the disease-associated genes.

SUMMARY OF THE INVENTION

[0009] Various embodiments of the invention herein include a conditionally immortal hematopoietic cell system for simulating at least one disease state, the system including: a population of CD34+ cells expressing at least one anti-apoptotic gene controlled by an inducible promoter; and at least one exogenous biomolecule to modulate expression of at least one disease-related gene selected from: RPS19, RPS14, RPL11, RPS24, RPS17, RPL35A, RPL5, RPS7, RPS10, RPS26, RPL26, RPL15, RPS29, TSR2, RPS28, RPL27, RPS27, GATA1, and a combination thereof. In certain embodiments, the at least one anti-apoptotic gene includes B-cell lymphoma extra-large (BCL-XL), c-Myc (MYC), or a combination thereof. In certain embodiments, the at least one disease-related gene is selected from: RPS19, RPS14, RPL11, and a combination thereof. In certain embodiments, the cell sytem further includes at least one exogenous biomolecule to modulate the activity or expression of p53. In certain embodiments, the at least one exogenous biomolecule includes a short-hairpin RNA (shRNA), a self-delivering RNA (sdRNA), or a combination thereof. In certain embodiments, the at least one exogenous biomolecule includes a CRISPR system. In certain embodiments, the exogenous biomolecule is in the form of a vector or plasmid which facilitates the formation of a final exogenous biomolecule in the target cell system which modulates the expression of at least one disease- related gene. In certain embodiments, the vector is a lentiviral vector. In certain embodiments, the activity of the exogenous biomolecule is controlled by an inducible promoter. In certain embodiments, the exogenous biomolecule is constitutively active.

[0010] In certain embodiments, the cell system includes a population of CD34+ cells expressing B-cell lymphoma extra-large (BCL-XL) and c-Myc (MYC) controlled by an inducible promoter; and at least one shRNA or sdRNA to modulate expression of at least one disease-related gene selected from: RPS19, RPS 14, RPL11, and a combination thereof. In certain embodiments, the cell system includes a population of CD34+ cells expressing B-cell lymphoma extra-large (BCL-XL) and c-Myc (MYC) controlled by an inducible promoter; and at least one CRISPR system to modulate expression of at least one disease-related gene selected from: RPS19, RPS 14, RPL11, and a combination thereof.

[0011] In certain embodiments, the population of CD34+ cells are from a healthy subject. In certain embodiments, the population of CD34+ cells are from a diseased subject. In certain embodiments, the population of CD34+ cells are hematopoietic stem cells. In certain embodiments, the population of CD34+ cells originate from a source selected from the group consisting of: umbilical cord blood, peripheral blood frozen stock, fresh peripheral blood, and a combination thereof. In certain embodiments, the disease state is Diamond-Blackfan anemia (DBA). In certain embodiments, the disease state is 5q- myelodysplasia. In certain

embodiments, the population of CD34+ cells further comprise a constitutive knockdown of at least one gene selected from the group consisting of: RPS19, RPS14, RPLl l, and a combination thereof.

[0012] Various embodiments of an invention herein provide a method of preparing the conditionally immortal cell system for simulating at least one disease state, the method including: introducing into a population of CD34+ cells at least one anti-apoptotic gene under the control of a promoter that is inducible by contacting the cells with a factor, wherein removing the factor from the cells reduces expression of the at least one anti-apoptotic gene; and introducing into a population of CD34+ cells at least one exogenous biomolecule to modulate the level of expression of at least one disease-related gene selected from: RPS 19, RPS14, RPLl l, RPS24, RPS 17, RPL35A, RPL5, RPS7, RPS 10, RPS26, RPL26, RPL15, RPS29, TSR2, RPS28, RPL27, RPS27, GATA1, and a combination thereof thereby simulating the disease state in the conditionally immortal cell system. In certain embodiments, the at least one disease- related gene is selected from: RPS19, RPS14, RPL11, and a combination thereof. In certain embodiments, the at least one anti-apoptotic gene is delivered using a lentiviral vector.

[0013] In certain embodiments, the method further includes introducing into the population of CD34+ cells at least one exogenous biomolecule to modulate the activity or expression of p53. In certain embodiments, at least one exogenous biomolecule includes a short-hairpin RNA (shRNA), a self-delivering RNA (sdRNA), or a combination thereof. In certain embodiments, at least one exogenous biomolecule includes a CRISPR system. In certain embodiments, the exogenous biomolecule is in the form of a vector or plasmid which facilitates the formation of a final exogenous biomolecule in the target cell system which modulates the expression of at least one disease-related gene. In certain embodiments, the expression of the exogenous biomolecule is controlled by an inducible promoter. In certain embodiments, the exogenous biomolecule is constitutively expressed. In certain embodiments, the method further includes the step of transducing the population of CD34+ cells with a construct encoding a short-hairpin RNA (shRNA) to knockdown at least one disease-related gene selected from the group consisting of: RPS 19, RPS14, RPL11, and a combination thereof. In certain embodiments, the method further includes contacting the population of CD34+ cells with a self-delivering RNA (sdRNA) to knockdown at least one disease-related gene selected from the group consisting of: RPS19, RPS14, RPL11, and a combination thereof. In an aspect of the invention, the factor may be an antibiotic. For example, the antibiotic is doxycycline. In certain embodiments, the anti-apoptotic gene is B-cell lymphoma extra-large (BCL-XL) or Myc (MYC), or a combination thereof. In certain embodiments, the disease state is Diamond-Blackfan anemia. In certain embodiments, the disease state is 5q- myelodysplasia. In certain embodiments, the method further includes removing the factor from the population of CD34+, thereby differentiating of the population of CD34+ cells. For example, the differentiating results in a greater number of erythroid cells than the population of CD34+ cells without introducing at least one anti-apoptotic gene.

[0014] Various embodiments of the invention herein provide a method of identifying compounds capable of rescuing, affecting or treating at least one disease state in a disease- simulating cell system of the present disclosure, the method including the steps of: providing a cell system simulating at least one disease state, as described herein; contacting the cell system with a compound; and characterizing the effect of the compound on the cell system contacted with the compound. In certain preferred embodiments, the cell system is a conditionally immortal cell system.

[0015] In certain embodiments, the step of characterizing the effect of the compound on the cell system comprises characterizing the levels of altered gene expression (from contacting a compound with the cell system) of one or more disease-related genes in the cell system. In certain embodiments, the characterizing step further comprises comparing the altered gene expression of the disease-related gene in the in the disease-simulating cell system with the corresponding level of gene expression in a cell system of a normal state. In certain

embodiments, gene expression is characterized using RNA-seq. In certain embodiments, the at least one disease-related gene is selected from RPS 19, RPS14, RPL11, RPS24, RPS17, RPL35A, RPL5, RPS7, RPSIO, RPS26, RPL26, RPL15, RPS29, TSR2, RPS28, RPL27, RPS27, GATA1, and a combination thereof. In certain embodiments, the at least one disease-related gene is selected from P53, BAX, P21, GADD45A, CDK 1A, BAG1, MDM2, and a combination thereof. In certain embodiments, the least one disease-related gene is selected from the group consisting of: RPS19, RPS14, RPL11, and a combination thereof.

[0016] In certain embodiments, the step of characterizing the effect of the compound on the cell system comprises characterizing the altered binding profile (from contacting a compound with the cell system) of one or more genomic signaling centers in at least one insulated neighborhood comprising a disease-related gene. In certain embodiments, the characterizing step further comprises comparing the altered GSC binding profile in the disease-simulating cell system with the corresponding GSC binding profile in a cell system of a normal state. In certain embodiments, GSC binding profile is determined using ChlP-seq. In certain embodiments, the at least one disease-related gene is selected from RPS 19, RPS14, RPL11, RPS24, RPS17, RPL35A, RPL5, RPS7, RPSIO, RPS26, RPL26, RPL15, RPS29, TSR2, RPS28, RPL27, RPS27, GATA1, and a combination thereof. In certain embodiments, the at least one disease-related gene is selected from P53, BAX, P21, GADD45A, CDK 1A, BAG1, MDM2, and a combination thereof. In certain embodiments, the least one disease-related gene is selected from the group consisting of: RPS19, RPS14, RPL11, and a combination thereof.

[0017] In certain embodiments, the compound is at least one compound selected from nutlin 3, pifithrin-a (PFT-a), nutlin 3a, and cyclic pifithrin-a. In certain embodiments, the method further includes calculating an optimal concentration of the compound for rescuing the at least one disease state. In certain embodiments, the method further includes calculating a half maximal effective concentration (EC50). In certain embodiments, the method further includes repeating the method steps one or more times to screen a plurality of compounds. In certain embodiments, the disease state is Diamond-Blackfan anemia. In certain embodiments, the disease state is 5q- myelodysplasia.

BRIEF DESCRIPTION OF THE DRAWINGS

[0018] FIG. 1 is an illustration comparing the number of cells throughout expansion and differentiation of DBA patient cells (dash line), the healthy patient cells (solid line), and the conditionally immortalized inducible DBA phenotype model disease model (dotted line).

[0019] FIGs. 2A-3C are gene tracks of ChlP-seq results comparing the binding profiles during differentiation and expansion of CD34+ progenitor cells. FIG. 2A shows the insulated neighborhood including apoptosis-associated gene GADD45A. FIG. 2B shows the insulated neighborhood including p53 pathway-associated gene RPS19. FIG. 2C shows the insulated neighborhood including CDKN1A.

[0020] FIGs. 3A-3D show the loglO p-values for genomic annotation of the binding sites of p53 in hematopoietic stem cells (HSCs) and erythroid cells (loglO p-values < 4 were considered significant). FIG. 3 A characterizes p53 binding sites in hematopoietic stem cell promoter regions. FIG. 3B characterizes p53 binding sites in erythroid cell promoter regions. FIG. 3C characterizes p53 binding sites in non-promoter regions of hematopoietic stem cells. FIG. 3D characterizes p53 binding sites in non-promoter regions of erythroid cells.

DETAILED DESCRIPTION OF THE INVENTION

I. INTRODUCTION

[0021] Ribosomal protein mutations have been implicated in the pathophysiology of DBA. The first gene, mutated in approximately 25% of DBA patients, was identified as RPS19 (ribosomal protein S19) (Gustavsson et al., Nat Genet. 1997 Aug;16(4):368-71 ; Draptchinskaia et al, Nat Genet. 1999 Feb;21(2): 169-75). Sequencing of patient samples has identified mutations of either large (60s) or small (40s) subunit ribosomal proteins in over 50% of patients (Vlachos et al, Br J Haematol. 2008 Sep; 142(6): 859-876). Identified genes include but are not limited to RPS19, RPL5, RPS10, RPL11, RPL35A, RPS7, RPS17, RPS24, RPL26, RPS26 and GATA1 genes, and most recently RPS29 (Mirabello et al, Blood. 2014 Jul 3;124(l):24-32). Some mutations of unknown significance are reported in other ribosomal protein genes (Doherty et al, Am J Hum Genet 2010;86(2):222-8). Patients are heterozygous for these mutations, always maintaining a wildtype copy of the affected RP gene. However, approximately 30% of people with DBA have no detectable RP mutation. Some phenotype/genotype correlations are known, relating to congenital abnormalities (Gazda et al, Am J Hum Genet. 2008;83(6):769-80; Quarello et al, Haematologica. 2010;95(2):206-13).

[0022] Diamond-Blackfan anemia-1 (DBA1, OMIM #105650) is caused by heterozygous mutations in the RPSl 9 gene on chromosome 19ql3. Other forms of DBA include DBA2 (OMIM #606129), caused by mutations on chromosome 8p23-p22; DBA3 (OMIM #610629), caused by mutation in the RPS24 gene on 10q22; DBA4 (OMIM #612527), caused by mutation in the RPSl 7 gene on 15q; DBA5 (OMIM #612528), caused by mutation in the RPL35A gene on 3q29; DBA6 (OMIM #612561), caused by mutation in the RPL5 gene on lp22.1; DBA7 (OMIM #612562), caused by mutation in the RPL11 gene on lp36; DBA8 (OMIM #612563), caused by mutation in the RPS7 gene on 2p25; DBA9 (OMIM #613308), caused by mutation in the RPS10 gene on 6p; DBA10 (OMIM #613309), caused by mutation in the RPS26 gene on 12q; DBA11 (OMIM #614900), caused by mutation in the RPL26 gene on 17pl3; DBA12 (OMIM #615550), caused by mutation in the RPL15 gene on 3p24; DBA13 (OMIM #615909), caused by mutation in the RPS29 gene on 14q; DBA14 (OMIM #300946), caused by mutation in the TSR2 gene on Xpl 1 ; DBA15 (OMIM #606164), caused by mutation in the RPS28 gene on 19pl3; DBA16 (OMIM #617408), caused by mutation in the RPL27 gene on chromosome 17q21 ; and DBA17 (OMIM #617409), caused by mutation in the RPS27 gene on chromosome lq21.

[0023] About one-quarter of DBA patients have a variety of mutations in RPS 19. See Gazda et al, British Journal of Haematology, 127(1): 105-13 (2014), which is hereby incorporated by reference in its entirety. The main phenotype of the ribosomal gene haploinsufficiency causing DBA is a failure of erythropoiesis. See Narla et al, Blood, 115(16):3196-3205, (2010); Lipton et al, Current Opinion Pediatrics, 22(1): 12-19, (2010); Ganapathi et al, British Journal of Hematology, 141(3): 376-387, (2008); Liu et al, Blood, 107(12):4583-4588, (2006), which are hereby incorporated by reference in its entirety. Knockdown of RPS14 or RPS 19 has also been observed to induce the mRNA expression of well-established p53 target genes: p21 and BAX, and reduce the mRNA expression of GATA1 as shown by quantitative RT-PCR (qRT-PCR). See, Dutt et al, Blood, 3;117(9):2567-76 (2011), which is hereby incorporated by reference in its entirety. Activation of p53 was demonstrated to be sufficient to impair erythropoiesis. [0024] Cord blood (CB) or peripheral blood (PB) are valuable sources of CD34⁺ cells and a means for ex vivo expansion or in vitro generation of erythrocytes for research of disease including DBA, but the inability to produce enough numbers of progenitor and differentiated cells for analysis in next generation sequencing techniques, such as ChlP-seq and HiChIP, remained a challenge. Individual blood progenitor cells in the hematopoietic systems have their own proliferation program and have the potential to self-replicate. Therefore, identification of the self-repli cation factors in lineage blood progenitors enabled immortalization of the cells. See, Hirose et al, Stem Cell Reports, l(6):499-508 (2013), which is hereby incorporated by reference in its entirety. Transduction of c-MYC and BCL-XL into multipotent hematopoietic progenitor cells derived from pluripotent stem cells enabled sustained exponential self- replication of glycophorin A+ erythroblasts. A doxycycline (Dox)-inducible system was used in the immortalized cell system to turn genes on or off.

[0025] To model haploinsufficiency herein, short hairpin RNAs (shRNA) were utilized in a conditionally immortalized hematopoietic cell system to decrease the expression of the RPS 14, RPL11, p53, or RPS19 genes by 40% to 60% at mRNA and protein levels. Knockdown of RPS19 was observed to increase the expression of the well-established p53 target genes, p21 and BAX as measured by quantitative real time PCR (qRT-PCR). The mRNA expression of p21 and BAX was observed to increase relative to β-actin following knockdown of RPS 19. The induction of p53 caused by RPS 14 or RPS19 knockdown approximated the levels of p53 induced by gamma irradiation, a positive control for p53 induction. To determine if the accumulated p53 protein was transcriptionally active, the expression of p53 target genes was examined.

[0026] This strategy allows one to maintain primary cells in culture for extended periods of time allowing an increase in biomass for experimentation, and provides the option to genetically engineer these cells to model disease and use these cells as a screening platform for additional compounds. Such a platform provides versatility to study red cell biology and may obviate the need to purchase primary cells. At present, attained DBA material is largely limited to RNA- based analysis due to low biomass (cells have growth/differentiation defect). Using the approaches proposed herein, there is a high likelihood of achieving the cell numbers (biomass) necessary to perform essential downstream assays to decipher the signaling centers and genome architecture in these cells. Further, developing an in vitro assay also leads to a better understanding of the disease, potential patient stratification methods, better clinical trial design, and a screening platform of disease model most closely related to in vivo target.

[0027] Additionally, other related dieases, such as 5q- myelodysplasia, may also be modeled with the cell system described herein. 5q- myelodysplasia (also known as Del 5q, 5q- syndrome, chromosome 5q deletion syndrome, or chromosome 5q monosomy) is a rare form of myelodysplasia syndrome. It is caused by deletion of a region of DNA in the long arm (q arm, band 5q31.1) of human chromosome 5, which contains 40 genes including t e RPS14 locus (Ebert et al., Nature. 2008 Jan 17;451 (7176):335-9). The loss of the RPS14 gene leads to the problems with red blood cell development characteristic of 5q- myelodysplasia. In patients with 5q- myelodysplasia, induction of p53 and up-regulation of the p53 pathway was also observed (Pellagatti et al, Blood. 2010 Apr l ; 115(13):2721 -3).

[0028] Described herein are compositions and methods for modeling diseases which allow assessment of efficacy of compounds in treating the disease. The effects of compounds on genomic signaling centers associated with the disease-related genes are also examined. The details of one or more embodiments of the invention are set forth in the accompanying description below. Although any materials and methods similar or equivalent to those described herein can be used in the practice or testing of the present invention, the preferred materials and methods are now described. Other features, objects and advantages of the invention will be apparent from the description. In the description, the singular forms also include the plural unless the context clearly dictates otherwise. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. In the case of conflict, the present description will control.

[0029] The present invention is further illustrated by the following non-limiting examples. II. DEFINITIONS

[0030] The definitions herein are non-limiting.

[0031] The term "disease-related gene", as used herein, refers to genes, either protein-coding or non-protein coding, whose activity and/or expression levels are altered in the disease phenotype.

[0032] The term, "factor", as used herein, refers to a compound or ligand that interacts with an inducible promoter of a vector to increase or decrease activity of the inducible promoter.

[0033] The terms "insulated neighborhood" and "IN", as used herein, refer to a region of a genome which includes a loop structure formed by two interacting sites in a chromosome sequence. The term "neighborhood gene" refers to a gene localized within an insulated neghborhood. IN architecture is generally defined by at least two interacting cites which come together, directly or indirectly, to form a DNA loop. These interacting sites can include CCCTC- Binding factor (CTCF) and are often co-occupied by cohesin. The integrity of these cohesin- associated interacting structures can affect the expression of neighborhood genes within the IN, as well as those genes in the vicinity of the INs. [0034] The term, "knockdown", as used herein, refers to a reduction of expression of a gene.

[0035] The term "modulate expression", as used herein, refers to a change in expression of a gene resulting from administration of an exogenous biomolecule.

[0036] The terms "genomic signaling center", and "GSC", as used herein, refer to a defined region of a genome which interacts with a defined set of biomolecules, such as transcription factors or other signaling molecules, to regulate gene expression in a context-specific manner. Genomic signaling centers can include enhancers bound by a highly context-specific combinatorial assemblies of transcription factors, signaling molecules, and chromatin remodeling proteins. These molecules are recruited to the site through cellular signaling.

Genomic signaling centers include multiple molecules that interact to form a three-dimensional transcription factor hub macrocomplex. Signaling centers are generally associated with one to four genes in a loop organized by biological function.

[0037] The term "promoter" as used herein, refers to a DNA sequence that defines where transcription of a gene by RNA polymerase begins and defines the direction of transcription indicating which DNA strand will be transcribed.

[0038] The term, "self-delivering RNA" or "sdRNA", as used herein, refers to an artificial chemically-modified RNA molecule that does not require a delivery vehicle for administration to a cell.

[0039] The term, "short-hairpin RNA" or "shRNA", as used herein, refers to an artificial RNA molecule having a tight hairpin turn that can interfere with gene expression when delivered to a cell. shRNAs can be delivered to a cell by a vector or plasmid for transcription by the cell.

[0040] The terms "clustered regulatory interspaced short palindromic repeat system" and "CRISPR system" refer to a system of biomolecules for modulating the expression of a target gene which includes a Guide RNA component (gRNA or sgRNA) and a CRISPR-associated endonuclease protein (CAS protein), most commonly a CAS -9 protein.

[0041] The term, "therapeutic target", as used herein, is a region of the genome that includes a genomic signaling center or encodes a signaling protein such as a transcription factor or a chromatin remodeling protein or is a signaling protein that is associated with modulation of expression of disease-associated genes.

[0042] The term "binding profile", as used herein, refers to the context-specific combination of biomolecules or factors identified to bind a signaling center, such as a GSC, that interact to form a three-dimensional macrocomplex. Factors may include master transcription factors, signaling transcription factors, chromatin remodelers, and the like. [0043] Described herein are compositions and methods for modeling diseases with hematopoietic cell systems. The details of one or more embodiments of the invention are set forth in the accompanying description below. Although any materials and methods similar or equivalent to those described herein can be used in the practice or testing of the present invention, the preferred materials and methods are now described. Other features, objects and advantages of the invention will be apparent from the description. In the description, the singular forms also include the plural unless the context clearly dictates otherwise. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. In the case of conflict, the present description will control.

[0044] The present invention is further illustrated by the following non-limiting examples.

III. EXAMPLES

Example 1. Experimental procedures

A. Erythroid cell culture

[0045] CD34⁺ cells were isolated from cord blood and peripheral blood (about lxlO⁶). Cells were expanded over 1 1 days to collect a total of 100x10⁶ cord blood cells and 25x10⁶ peripheral blood cells. The cells were conditionally immortalized as described below on the third day, differentiated in erythroid medium and collected on the seventh day (about 400xl0⁶ cord blood CD34+ cells or about l OOxlO⁶ peripheral blood CD34+ cells).

B. Media composition

[0046] The thaw medium contained 6mL isotonic percoll and 14mL high glucose DMEM (Invitrogen # 11965 or similar). The plating medium contained lOOmL Williams E medium (Invitrogen #A1217601, without phenol red) and the supplement pack #CM3000 from

ThermoFisher Plating medium containing 5mL FBS, Ι ΟμΙ dexamethasone, and 3.6mL plating/maintenance cocktail. Stock trypan blue (0.4%, Invitrogen # 15250) was diluted 1 :5 in PBS.

[0047] The ThermoFisher complete maintenance medium contained supplement pack #CM4000 (Ι μΐ dexamethasone and 4mL maintenance cocktail) and lOOmL Williams E

(Invitrogen #A1217601, without phenol red).

[0048] The modified maintenance media had no stimulating factors (dexamethasone, insulin, etc.), and contained lOOmL Williams E (Invitrogen #A1217601, without phenol red), lmL L- Glutamine (Sigma #G7513) to 2mM, 1.5mL HEPES (VWR #J848) to 15mM, and 0.5mL penicillin/streptomycin (Invitrogen # 15140) to a final concentration of 50U/mL each. C. DNA purification

[0049] DNA purification was conducted as described in Ji et al, PNAS 112(12):3841-3846 (2015) Supporting Information, which is hereby incorporated by reference in its entirety. One milliliter of 2.5 M glycine was added to each plate of fixed cells and incubated for 5 minutes to quench the formaldehyde. The cells were washed twice with PBS. The cells were pelleted at 1,300 g for 5 minutes at 4°C. Then, 4 ^χ 10⁷ cells were collected in each tube. The cells were lysed gently with 1 mL of ice-cold Nonidet P-40 lysis buffer containing protease inhibitor on ice for 5 minutes (buffer recipes are provided below). The cell lysate was layered on top of 2.5 volumes of sucrose cushion made up of 24% (wt vol) sucrose in Nonidet P-40 lysis buffer. This sample was centrifuged at 18,000 g for 10 minutes at 4°C to isolate the nuclei pellet (the supernatant represented the cytoplasmic fraction). The nuclei pellet was washed once with PBS/1 mM EDTA. The nuclei pellet was resuspended gently with 0.5mL glycerol buffer followed by incubation for 2 minutes on ice with an equal volume of nuclei lysis buffer. The sample was centrifuged at 16,000 g for 2 minutes at 4°C to isolate the chromatin pellet (the supernatant represented the nuclear soluble fraction). The chromatin pellet was washed twice with PBS/1 mM EDTA. The chromatin pellet was stored at -80 °C.

[0050] The Nonidet P-40 lysis buffer contained 10 mM Tris HCl (pH 7.5), 150 mM NaCl, and 0.05% Nonidet P-40. The glycerol buffer contained 20 mM Tris HCl (pH 7.9), 75 mM NaCl, 0.5 mM EDTA, 0.85 mM DTT, and 50% (vol/vol) glycerol. The nuclei lysis buffer contained 10 mM Hepes (pH 7.6), 1 mM DTT, 7.5 mM MgC12, 0.2 mM EDTA, 0.3 M NaCl, 1 M urea, and 1% Nonidet P-40.

D. Chromatin immunoprecipitation sequencing (ChlP-seq)

[0051] ChlP-seq was performed using the following protocol for erythroid cells to determine the composition and confirm the location of genomic signaling centers.

i. Cell cross-linking

[0052] 2 x 10⁷ cells were used for each run of ChlP-seq. Two ml of fresh 11% formaldehyde (FA) solution was added to 20ml media on 15cm plates to reach a 1.1% final concentration. Plates were swirled briefly and incubated at room temperature (RT) for 15 minutes. At the end of incubation, the FA was quenched by adding 1ml of 2.5M Glycine to plates and incubating for 5 minutes at RT. The media was discarded to a 1L beaker, and cells were washed twice with 20ml ice-cold PBS. PBS (10ml) was added to plates, and cells were scraped off the plate. The cells were transferred to 15ml conical tubes, and the tubes were placed on ice. Plates were washed with an additional 4ml of PBS and combined with cells in 15ml tubes. Tubes were centrifuged for 5 minutes at 1,500 rpm at 4°C in a tabletop centrifuge. PBS was aspirated, and the cells were flash frozen in liquid nitrogen. Pellets were stored at -80°C until ready to use. ii. Pre-block magnetic beads

[0053] Thirty μΐ Protein G beads (per reaction) were added to a 1.5ml Protein LoBind Eppendorf tube. The beads were collected by magnet separation at RT for 30 seconds. Beads were washed 3 times with 1ml of blocking solution by incubating beads on a rotator at 4°C for 10 minutes and collecting the beads with the magnet. Five μg of an antibody was added to the 250μ1 of beads in block solution. The mix was transferred to a clean tube, and rotated O/N at 4°C. On the next day, buffer containing antibodies was removed, and beads were washed 3 times with 1.1ml blocking solution by incubating beads on a rotator at 4°C for 10 minutes and collecting the beads with the magnet. Beads were resuspended in 50μ1 of block solution and kept on ice until ready to use.

Hi. Cell lysis, genomic fragmentation, and chromatin immunoprecipitation

[0054] Complete® protease inhibitor cocktail was added to lysis buffer 1 (LB1) before use. One tablet was dissolved in 1ml of H2O for a 50x solution. The cocktail was stored in aliquots at -20°C. Cells were resuspended in each tube in 8ml of LB1 and incubated on a rotator at 4°C for 10 minutes. Nuclei were spun down at 1,350 g for 5 minutes at 4°C. LB1 was aspirated, and cells were resuspended in each tube in 8ml of LB2 and incubated on a rotator at 4°C for 10 minutes.

[0055] A Covaris^® E220evolution™ ultrasonicator was programmed per the manufacturer's recommendations for high cell numbers. Erythroid cells were sonicated for 12 minutes. Ly sates were transferred to clean 1.5ml Eppendorf tubes, and the tubes were centrifuged at 20,000 g for 10 minutes at 4°C to pellet debris. The supernatant was transferred to a 2ml Protein LoBind Eppendorf tube containing pre-blocked Protein G beads with pre-bound antibodies. Fifty μΐ of the supernatant was saved as input. Input material was kept at -80°C until ready to use. Tubes were rotated with beads overnight at 4°C.

iv. Wash, elution, and cross-link reversal

[0056] All washing steps were performed by rotating tubes for 5 minutes at 4°C. The beads were transferred to clean Protein LoBind Eppendorf tubes with every washing step. Beads were collected in 1.5ml Eppendorf tube using a magnet. Beads were washed twice with 1.1ml of sonication buffer. The magnetic stand was used to collect magnetic beads. Beads were washed twice with 1.1ml of wash buffer 2, and the magnetic stand was used again to collect magnetic beads. Beads were washed twice with 1.1ml of wash buffer 3. All residual Wash buffer 3 was removed, and beads were washed once with 1.1ml TE + 0.2% Triton X-100 buffer. Residual TE + 0.2% Triton X-100 buffer was removed, and beads were washed twice with TE buffer for 30 seconds each time. Residual TE buffer was removed, and beads were resuspended in 300μ1 of ChIP elution buffer. Two hundred fifty μΐ of ChIP elution buffer was added to 50μ1 of input, and the tubes were rotated with beads 1 hour at 65°C. Input sample was incubated overnight at 65°C oven without rotation. Tubes with beads were placed on a magnet, and the eluate was transferred to a fresh DNA LoBind Eppendorf tube. The eluate was incubated overnight at 65°C oven without rotation

v. Chromatin extraction and precipitation

[0057] Input and immunoprecipitant (IP) samples were transferred to fresh tubes, and 300μ1 of TE buffer was added to IP and Input samples to dilute SDS. RNase A (20mg/ml) was added to the tubes, and the tubes were incubated at 37°C for 30 minutes. Following incubation, 3μ1 of 1M CaC12 and 7μ1 of 20mg/ml Proteinase K were added, and incubated 1.5 hours at 55°C. MaXtract High Density 2ml gel tubes (Qiagen) were prepared by centrifugation at full speed for 30 seconds at RT. Six hundred μΐ of phenol/chloroform/isoamyl alcohol was added to each proteinase K reaction and transferred in about 1.2ml mixtures to the MaXtract tubes. Tubes were spun at 16,000 g for 5 minutes at RT. The aqueous phase was transferred to two clean DNA LoBind tubes (300μ1 in each tube), and 1.5μ1 glycogen, 30μ1 of 3M sodium acetate, and 900μ1 ethanol were added. The mixture was precipitated overnight at -20°C or for 1 hour at -80°C, and spun down at maximum speed for 20 minutes at 4°C. The ethanol was removed, and pellets were washed with 1ml of 75% ethanol by spinning tubes down at maximum speed for 5 minutes at 4°C. Remnants of ethanol were removed, and pellets were dried for 5 min at RT. Twenty-five μΐ of H2O was added to each immunoprecipitant (IP) and input pellet, left standing for 5 minutes, and vortexed briefly. DNA from both tubes was combined to obtain 50μ1 of IP and 50μ1 of input DNA for each sample. One μΐ of this DNA was used to measure the amount of pulled down DNA using Qubit dsDNA HS assay (ThermoFisher, #Q32854). The total amount of

immunoprecipitated material ranged from several ng (for TFs) to several hundred ng (for chromatin modifications). Six μΐ of DNA was analyzed using qRT-PCR to determine enrichment. The DNA was diluted if necessary. If enrichment was satisfactory, the rest was used for library preparation for DNA sequencing.

vi. Library preparation for DNA sequencing

[0058] Libraries were prepared using NEBNext Ultra II DNA library prep kit for Illumina (NEB, #E7645) using NEBNext Multiplex Oligos for Illumina (NEB, #6609S) according to manufacturer's instructions with the following modifications. The remaining ChIP sample (about 43μ1) and ^g of input samples for library preparations were brought up the volume of 50μ1 before the End Repair portion of the protocol. End Repair reactions were run in a PCR machine with a heated lid in a 96-well semi-skirted PCR plate (ThermoFisher, #AB1400) sealed with adhesive plate seals (ThermoFisher, #AB0558) leaving at least one empty well in-between different samples. Undiluted adapters were used for input samples, 1 : 10 diluted adapters for 5- lOOng of ChIP material, and 1 :25 diluted adapters for less than 5ng of ChIP material. Ligation reactions were run in a PCR machine with the heated lid off. Adapter ligated DNA was transferred to clean DNA LoBind Eppendorf tubes, and the volume was brought to 96.5μ1 using

[0059] 200-600bp ChIP fragments were selected using SPRIselect magnetic beads (Beckman Coulter, #B23317). Thirty μΐ of the beads were added to 96.5μ1 of ChIP sample to bind fragments that are longer than 600 bp. The shorter fragments were transferred to a fresh DNA LoBind Eppendorf tube. Fifteen μΐ of beads were added to bind the DNA longer than 200bp, and beads were washed with DNA twice using freshly prepared 75% ethanol. DNA was eluted using 17μ1 of 0.1X TE buffer. About 15μ1 of DNA was collected.

[0060] Three μΐ of size-selected Input sample and all (15μ1) of the ChIP sample was used for PCR. The amount of size-selected DNA was measured using a Qubit dsDNA HS assay. PCR was run for 7 cycles of for Input and ChIP samples with about 5-1 Ong of size-selected DNA, and 12 cycles with less than 5 ng of size-selected DNA. One-half of the PCR product (25μ1) was purified with 22.5μ1 of AMPure XP beads (Beckman Coulter, #A63880) according to the manufacturer's instructions. PCR product was eluted with 17μ1 of 0.1X TE buffer, and the amount of PCR product was measured using Qubit dsDNA HS assay. An additional 4 cycles of PCR were run for the second half of samples with less than 5ng of PCR product, DNA was purified using 22.5μ1 of AMPure XP beads. The concentration was measured to determine whether there was an increased yield. Both halves were combined, and the volume was brought up to 50μ1 using H2O.

[0061] A second round of purifications of DNA was run using 45 μΐ of AMPure XP beads in 17μ1 of 0.1X TE, and the final yield was measured using Qubit dsDNA HS assay. This protocol produces from 20ng to lmg of PCR product. The quality of the libraries was verified by diluting Ιμΐ of each sample with H2O if necessary using the High Sensitivity BioAnalyzer DNA kit (Agilent, #5067-4626) based on manufacturer's recommendations.

vii. Reagents

[0062] 11% Formaldehyde Solution (50mL) contained 14.9ml of 37% formaldehyde (final cone. 11%), 1 ml of 5M NaCl (final cone. 0.1 M), ΙΟΟμΙ of 0.5M EDTA (pH 8) (final cone. lmM), 50μ1 of 0.5M EGTA (pH 8) (final cone. 0.5mM), and 2.5 ml 1M Hepes (pH 7.5) (final cone. 50 mM).

[0063] Block Solution contained 0.5% BSA (w/v) in PBS and 500mg BSA in 100ml PBS. Block solution may be prepared up to about 4 days prior to use.

[0064] Lysis buffer 1 (LB1) (500ml) contained 25ml of 1 M Hepes-KOH, pH 7.5; 14ml of 5M NaCl; 1 ml of 0.5M EDTA, pH 8.0; 50ml of 100% Glycerol solution; 25ml of 10% NP-40; and 12.5ml of 10% Triton X-100. The pH was adjusted to 7.5. The buffer was sterile-filtered, and stored at 4 °C. The pH was re-checked immediately prior to use.

[0065] Lysis buffer 2 (LB2) (1000ml) contained 10ml of 1 M Tris-HCL, pH 8.0; 40ml of 5 M NaCl; 2ml of 0.5M EDTA, pH 8.0; and 2ml of 0.5M EGTA, pH 8.0. The pH was adjusted to 8.0. The buffer was sterile-filtered, and stored at 4 °C. The pH was re-checked immediately prior to use.

[0066] Sonication buffer (500ml) contained 25ml of 1M Hepes-KOH, pH 7.5; 14ml of 5M NaCl; 1ml of 0.5M EDTA, pH 8.0; 50ml of 10% Triton X-100; 10ml of 5% Na-deoxycholate; and 5ml of 10% SDS. The pH was adjusted to 7.5. The buffer was sterile-filtered, and stored at 4 °C. The pH was re-checked immediately prior to use.

[0067] Proteinase inhibitors were included in the LB1, LB2, and Sonication buffer.

[0068] Wash Buffer 2 (500ml) contained 25ml of 1M Hepes-KOH, pH 7.5; 35 ml of 5M NaCl; 1ml of 0.5M EDTA, pH 8.0; 50ml of 10% Triton X-100; 10ml of 5% Na-deoxycholate; and 5ml of 10% SDS. The pH was adjusted to 7.5. The buffer was sterile-filtered, and stored at 4 °C. The pH was re-checked immediately prior to use.

[0069] Wash Buffer 3 (500ml) contained 10ml of 1M Tris-HCL, pH 8.0; 1ml of 0.5M EDTA, pH 8.0; 125ml of 1M LiCl solution; 25ml of 10% NP-40; and 50ml of 5% Na- deoxycholate. The pH was adjusted to 8.0. The buffer was sterile-filtered, and stored at 4 °C. The pH was re-checked immediately prior to use.

[0070] ChIP elution Buffer (500ml) contained 25ml of 1 M Tris-HCL, pH 8.0; 10ml of 0.5M EDTA, pH 8.0; 50ml of 10% SDS; and 415ml of ddH20. The pH was adjusted to 7.5. The buffer was sterile-filtered, and stored at 4 °C. The pH was re-checked immediately prior to use.

E. Analysis of ChlP-seq results

[0071] All obtained reads from each sample were trimmed using trim galore 0.4.1 requiring a Phred score > 20 and a read length > 30. The trimmed reads were mapped against the human genome (hgl9 build) using Bowtie (version 1.1.2) with the parameters: -v 2 -m 1 -S -t. All unmapped reads, non-uniquely mapped reads and PCR duplicates were removed. All the ChlP- seq peaks were identified using MACS2 with the parameters: -q 0.01— SPMR. The ChlP-seq signal was visualized in the UCSC genome browser. ChlP-seq peaks that are at least 2 kb away from annotated promoters (RefSeq, Ensemble and UCSC Known Gene databases combined) were selected as distal ChlP-seq peaks.

F. RNA-seq

[0072] This protocol is a modified version of the following protocols: MagMAX ra^'rVana Total RNA Isolation Kit User Guide (Applied Biosystems #MAN0011131 Rev B.0), NEBNext Poly(A) mRNA Magnetic Isolation Module (E7490), and NEBNext Ultra Directional RNA Library Prep Kit for Illumina (E7420) (New England Biosystems #E74901).

[0073] The MagMAX mirW ana kit instructions (the section titled "Isolate RNA from cells" on pages 14-17) were used for isolation of total RNA from cells in culture. Two hundred μΐ of Lysis Binding Mix was used per well of the multiwell plate containing adherent cells (usually a 24-well plate).

[0074] For mRNA isolation and library prep, the NEBNext Poly(A) mRNA Magnetic Isolation Module and Directional Prep kit was used. RNA isolated from cells above was quantified, and prepared in 50C^g of each sample in 50μ1 of nuclease-free water. This protocol may be run in microfuge tubes or in a 96-well plate.

[0075] The 80% ethanol was prepared fresh, and all elutions are done in 0. IX TE Buffer. For steps requiring Ampure XP beads, beads were at room temperature before use. Sample volumes were measured first and beads were pipetted. Section 1.9B (not 1.9A) was used for NEBNext Multiplex Oligos for Illumina (#E6609). Before starting the PCR enrichment, cDNA was quantified using the Qubit (DNA High Sensitivity Kit, ThermoFisher #Q32854). The PCR reaction was run for 12 cycles.

[0076] After purification of the PCR Reaction (Step 1.10), the libraries were quantified using the Qubit DNA High Sensitivity Kit. Ιμΐ of each sample were diluted to 1-2¾/μ1 to run on the Bioanalyzer (DNA High Sensitivity Kit, Agilent # 5067-4626). If Bioanalyzer peaks were not clean (one narrow peak around 300bp), the AMPure XP bead cleanup step was repeated using a 0.9X or 1.0X beads:sample ratio. Then, the samples were quantified again with the Qubit, and run again on the Bioanalyzer (l-2ng^l).

[0077] Nuclear RNA from INTACT-purified nuclei or whole neocortical nuclei was converted to cDNA and amplified with the Nugen Ovation RNA-seq System V2. Libraries were sequenced using the Illumina HiSeq 2500.

G. RNA-seq data analysis

[0078] All obtained reads from each sample were mapped against the human genome (hgl9 build) using STAR_2.5.2b, which allows mapping across splice sites by reads segmentation (Dobin et al, Bioinformatics (2012) 29 (1): 15-21, which is hereby incorporated by reference in its entirety). The uniquely mapped reads were subsequently assembled into transcripts guided by reference annotation (RefSeq gene models) (Pruitt et al, Nucleic Acids Res. 2012 Jan;

40(Database issue): D130-D135, which is incorporated by reference in its entirety) with Cuffnorm v2.2.1 (Trapnell et al., Nature Protocols 7, 562-578 (2012), which is hereby incorporated by reference in its entirety). The expression level of each gene was quantified with normalized FPKM (fragments per kilobase of exon per million mapped fragments). The differentially expressed genes were called using Cuffdiff v2.2.1 with q value < 0.01 and log2fold change >=1 or <= -1.

H. qRT-PCR

[0079] qRT-PCR was performed as described in North et al, PNAS, 107(40) 17315-17320 (2010), which is hereby incorporated by reference in its entirety. qRT-PCR was performed with cDNA using the iQ5 Multicolor rtPCR Detection system from BioRad with 60°C annealing.

[0080] Analysis of the fold changes in expression as measured by qRT-PCR were performed using the technique below. The control was DMSO, and the treatment was the selected compound (CPD). The internal control was GAPDH or B-Actin, and the gene of interest is the target. First, the averages of the 4 conditions were calculated for normalization:

DMSO: GAPDH, DMSO:Target, CPD: GAPDH, and CPD:Target. Next, the ACT of both control and treatment were calculated to normalize to internal control (GAPDH) using

(DMSO:Target) - (DMSO:GAPDH) = ACT control and (CPD:Target) - (CPD: GAPDH) = ACT experimental. Then, the AACT was calculated by ACT experimental - ACT control. The Expression Fold Change was calculated by 2-( AACT) (2 -fold expression change was shown by RNA-Seq results provided herein).

I. Conditional immortalization of hematopoietic cells

[0081] Conditional immortalization of hematopoietic cells was performed as described in Hirose et al, 2013, Stem Cell Reports, Vol. 1 :499-508, the contents of which are hereby incorporated by reference in their entirety. For the immortalization of hematopoietic cells, human hematopoietic progenitor (CD34+) cells were purchased from commercial vendors as purified cells or isolated from peripheral blood. Hematopoietic progenitor cells, were collected on day 11 of culture, and 5 x 10⁴ cells were suspended in erythroid differentiation medium (Stemcell Technologies, Canada). Then, the cells were transduced with viral supernatant for DOX-inducible c-MYC, or BCL-XL, or a combination of c-MYC and BCL-XL (multiplicity of infection: MOI = 20), E6/E7, hTERT, or SV40T system. Viral transduction was achieved using the retronectin-transduction protocol (Clontech). Cells were cultured in the presence of ^g/ml DOX.

[0082] Erythroid cells were cultured in Erythroid differrntiation media supplemented with antibiotics blasticidin and/or puromycin and ^g/ml DOX. The cells were cultivated for an additional 12 days to assure all cells expressed the desired transgenes required for conditional immortalization.

J. Conditional immortalization of hepatocvtes

[0083] Cryopreserved hepatocytes were seeded on collagen-coated culture plates at a density of 30,000 cells/cm² in Hepatocyte growth medium (HGM) containing William's E basal medium supplemented with ΙΟΟμΜ dexamethasone, 3.75g/L Bovine Serum Albumin (BSA), l x insulin, transferrin and selenium (ITS), 1% penicillin-streptomycin, and 10 ng/mL oncostatin M (OSM) (Sigma-Aldrich, St. Louis, US). After 24 hours, the cells were transduced with viral particles containing HPV E6 and E7 genes. The cells were cultured for an additional 3 to 8 weeks in the presence of OSM. The medium was replaced every 2 to 3 days. Ten days after the transduction, proliferating colonies were observed in the transduced cell cultures. For stabilization of the epithelial phenotype of the cells, cultures were treated with 2 ίοΙΟμΜ of MEK1/2 inhibitor U0126 (Sigma-Aldrich, St. Louis, US) for 1 to 3 weeks. Colonies of proliferating hepatocytes maintaining their epithelial phenotype were selected by trypsinization and re-seeded at a density of 20,000 cells/cm² and sub-cultured. The number of population doublings (PD) was calculated at each passage. The expression of CYP450, HPV E6, and HPV E7 was evaluated at population doubling number 19. When the colonies were expanded to 120 x 10⁶ cells, the hepatocytes were cryopreserved in HGM containing 20% FBS and 10% DMSO (See, Levy et al, 2015, Nat Biotechnol;33(12): 1264-1271, the contents of which are hereby incorporated by reference in their entireties).

K. Generating stable knockdown cell lines

[0084] 293T cells were transfected over 4 days using pLKO (constitutive, U6 promoter), pTPJPZ (dox inducible, minimal cytomegalovirus, red fluorescent protein [RFP]), or pSMART (dox inducible, EF1 alpha RFP, or phosphogly cerate kinase green fluorescent protein [PGK GFP]) plasmids to generate shRNA lentiviruses.

[0085] Conditionally immortalized cord blood CD34+ cells were transduced with the shRNA lentiviruses using the retronectin-transduction protocol (Clontech). Puromycin selection was performed 24-48 hours post-transduction. Cells were counted to monitor cell viability. After 12 days, which allowed for a large biomass, knockdown efficiency was analyzed. qRT-PCR was used to identify changes in mRNA, and Western blot was used to determine levels of protein depletion.

[0086] Cryopreserved CD34+ cells (cord blood, peripheral blood, or bone marrow derived) were thawed from cryovials (typically 1 x 10⁶ or 5 x 10⁶ cells/vial) at 37°C until completely thawed. Cells were then grown to a density of 1 x 10⁶ cells/mL for a period of 11 days in Hematopoietic Stem Cell expansion media DXF (Promocell) for the expansion of hematopoietic progenitor cells. Expanded CD34+ progenitor cells were then grown for an additional 3 days for erythroid differentiation in SFEMII medium (Stem Cell Technologies) supplemented with the Erythroid Expansion supplement (Stem Cell Technologies). Cells can be efficiently

cryopreserved at this step and are ready for viral transduction when virus is prepared. [0087] The vector and the shRNA of interest were selected from the plasmid list of Table 1. The SEQ ID NOs represent the nucleic acid sequence targets of the shRNA.

Table 1. Plasmid list

[0088] Additional vectors that were used include those listed in Table 2. The table provides the Clone ID, Target, Modification, Tag, Inducible or Constitutive (I/C), and Selection method.

Table 2. Additional lasmids

pL VX-EF 1 a-IRES-bla-DEST Destination None EFla NA I Bla pLVX-EFla-IRES-hygro NA None EFla NA I Hygro pL VX-EF 1 a-IRES-hygro-DEST Destination None EFla NA I Hygro pLVX-EFla-IRES-neo NA None EFla NA I Neo pLVX-EFla-IRES-neo-DEST Destination None EFla NA I Neo pL VX-EF 1 a-IRES-puro NA None EFla NA I Puro pLVX-EFla-IRES-puro-DEST Destination None EFla NA I Puro pLVX-TetOne Myc Overexpression PGK NA I Puro pLVX-TetOne BCLXL Overexpression PGK NA I Blast pLVX-TetOne Myc Overexpression PGK NA I Blast pLVX-TetOne BCLXL Overexpression PGK NA I Puro pLVX-TetOne-Bla BCLXL/Myc IRES PGK NA I Blast pLVX-TetOne-Bla eGFP Control GFP NA Blast pLVX-TetOne-Bla BCLXL Myc T2A PGK NA NA Puro pLVX-TetOne-Bla empty Empty Control PGK NA NA Bla pLVX-TetOne-Bla-DEST Destination none PGK NA NA Bla pLVX-TetOne-Bla-HPV16 E6/E7 NA HP VI 6 E6/E7 PGK NA NA Bla pLVX-TetOne-Bla-hTERT NA hTERT PGK NA NA Bla pLVX-TetOne-Bla-SV40T NA SV40T PGK NA NA Bla pLVX-TetOne-Hygro NA none PGK NA NA Hygro pLVX-TetOne-Hygro-DEST Destination none PGK NA NA Hygro pLVX-TetOne-Puro NA none PGK NA NA Puro pLVX-TetOne-Puro-DEST Destination none PGK NA NA Puro pSMART RPS19 shRNA KD EF1 alpha RFP NA Puro

PSMART RPS19 shRNA KD EF1 alpha RFP NA Puro

PSMART RPS19 EFP-RFP shRNA KD EF1 alpha RFP NA Puro

PSMART NT shRNA KD EF1 alpha RFP NA Puro pSMART NT shRNA KD EF1 alpha GFP NA Puro

PSMART RPS19 1 shRNA KD pGK GFP NA Puro

PSMART RPS19 2 shRNA KD PGK GFP NA Puro

PSMART NT shRNA KD pGK GFP NA Puro pTRIPZ-Neo Empty NA CMV RFP NA Neo pTRIPZ-Puro-sh T NT NT CMV RFP NA Puro pTRIPZ -tGFP . Trad stuff er NA NA RFP NA NA

[0089] Table 3 provides additional inserts to be used in any of the plasmids listed in Tables 1 and 2.

Table 3. Additional lasmid inserts

[0090] In additional embodiments, RPS19_shl (Ctacgatgagaactggttct; SEQ ID NO: 45) is inserted into one of the plasmids listed in Tables 1 and 2. [0091] 293FT cells (Clontech) were transfected with 7μg of pLVX-TetOne-Puro lentiviral vector combined with Lenti-X Packaging Single Shots (VSV-G) plasmids (Clontech). 293FT cells were transfected using the plasmid mixture. For the transfection of 293FT cells with shRNA, 7μg of combined shRNA lentivector, VSV-G, and packaging plasmids were added to the cells using Fugene reagent and OPTIMEM media.

[0092] For both transgene and shRNA transfection, 48 hours after the transfection, the viral supernatant was collected and filtered through a 0.45uM PVDF unit. The viral particles were concentrated by adding 1/3 total volume of Lenti-X Concentrator (Clontech). The mixture was mixed well by inversion and incubated for 1-2 hours at 4°C. After precipitation, the tubes were centrifuged at 4°C for more than 45 minutes at 1,500 g to pellet the virus. Then, the supernatant was aspirated and the pellet was resuspended in PBS to a concentration of 25-50X. Plates were coated with retronectin and incubated for more than 4 hours at room temperature or overnight at 4°C, using the Retronectin-transduction protocol (Clontech) according to the manufacturer's instructions. Following retronectin incubation, the plates were blocked with 2% BSA/PBS for more than 30 minutes, and washed with PBS. Target cells were seeded onto the retronectin- coated plates and the appropriate amount of virus was added. Then the plates were spun at 2,500 rpm for 1-2 hours at RT. Cells were incubated at 37°C for 72 hours following the infection and prior to selection.

[0093] For selection, antibiotic was added in the cell culture medium (about 5μg/mL of PURO). Following the death of all the non-resistant cells, the surviving cells were collected for RNA and protein assay. Alternatively, cells were transduced with lentiviral vectors containing a GFP reporter (GFPtrad) expressed in the cell membrane. Cells expressing membrane bound GFP were magnetically enriched using a GFP specific antibody in conjuction with MACS cell separation (Miltenyi Biotec). This approach could not enrich cell population expressing intracellular GFP; however, it enriched for cells expressing GFP on the cell membrane.

L. Knockdown efficiency

[0094] For RNA extraction, 1 x 10⁶ cells were harvested in 500μί of TRIZOL reagent (Thermo Scientific), and incubated for 5 minutes at room temperature. Then, ΙΟΟμί of chloroform was added to the cells, and the mixture was vortexed and then centrifuged at 12,000x gravity for 15 minutes. The aqueous phase was transferred to fresh tubes and one volume of 70% ethanol was added. The TRIZOL Plus RNA purification Kit (Thermo Scientific) was used for the remaining extraction protocol according to the manufacturer's instructions. For the conversion of RNA to cDNA the High Capacity cDNA RT kit (Thermo Scientific) was used according to the manufacturer's instructions. The cDNA was analyzed with qRT-PCR to determine the fold change of knockdown cells using Taqman Fast PCR mix and Applied Biosystems probes (Thermo Scientific).

M. Determining the IC50

[0095] The optimal concentration prior to cell toxicity (IC50) was determined for each compound that mimics or rescues the DBA phenotype. First, the IC50 for 4 different compounds in a TF1 cell line and CB-CI-IRES-CD34+ primary cells were determined. Nutlin 3 is a potent and selective MDM2 (RING finger-dependent ubiquitin protein ligase for itself and p53) antagonist with an IC50 of 90nM in a cell-free assay. It has also been observed to stabilize p73 in p53-deficient cells. Pifithrin-a is an inhibitor of p53, which inhibits p53-dependent transactivation of p53-responsive genes. Nutlin 3a, an active enantiomer of Nutlin 3, inhibits the p53/MDM2 interaction with an IC50 of 90 nM in a cell-free assay. Cyclic Pifithrin-α was also analyzed. Three 96-well plates were used for analysis of each cell type.

[0096] Three rows were dedicated to each compound, and column 12 contained

cells+media/2% DMSO. Compound dilutions were made in 11 tubes containing media + 2% DMSO (11 x 5= 55 x 1.5 ml tubes per cell type). Two conical tubes of media+2% DMSO were prepared for each media type (RPMl-GMCSF or Diff media-dox). Seven ml of media was combined with each drug. Therefore, 35ml was needed for each cell type, and 50mL+2% DMSO media was made. The compounds were diluted to lOmM in DMSO. Cells were seeded at 100K cells/well in 50μί of media. Cells were diluted to 2M cells/ml. One hundred eight wells were needed for each cell type (72 wells on 2 plates, 36 wells on 1 plate).

[0097] For each drug concentration, 150μί of each dilution was analyzed in triplicate for each cell type. 0.5 ml of each drug concentration was made resulting in 300μί extra.

[0098] Drug concentrations analyzed were as shown in Table 4.

[0099] 0.5mL of DMSO-media was added to each of 55 1.5ml tubes. Twenty μΐ. of lOmM stock and 480μί of additional media was added into tube 1. 10.8 million cells were needed in 5.4ml aliquots for 108 wells with 100,000 cells/well in 50μ1. of liquid. Eight ml of cells at 2M/ml were collected for a total of 16 million cells. An additional 5.2 million cells were transferred by 2.6 ml aliquots to 2 sterile reservoirs (for 10ml of cells). The final titration calculation was as follows: 8ml of 2 million/cells = 16 million cells/8ml = 2,000 cells^tL x 50μΙ.= 100,000 cells^L/well. N. Biological activity of maximal compound dosages

[0100] The biological activity of compounds inn TF1 and CI-CB-CD34+ cells was studied by choosing a maximal dose that does not cause cell toxicity.

[0101] The concentration of 10μΜ was used for compounds Nutlin 3, TFP, Pifithrin-a, Nutlin 3a, and cyclic Pifithrin-a. The stock for each compound was at lOmM and were diluted to ΙΟΟμΜ with RMPI+GCMSF (2% DMSO). Cells were collected and resuspended in 8 mL of RPMI+GMCSF at a concentration of lxlO⁶ cells/mL. For the TF1 cells, at each well of a 6-well plate were added lmL of cells, 300μί of ΙΟΟμΜ compound dilution, and 1.7mL of

RMPI+GCMSF+1.5%DMSO (1% final DMSO concentration in 3mL total). One mL of the CI- CB-CD34+ cells, 300μί of ΙΟΟμΜ compound dilution, and 700μΙ. of RMPI+GCMSF (0.1% final DMSO concentration in 2mL total) were added to each well of a 6-well plate.

O. Dose titration of compounds into RPS19 KD TF1 cells

[0102] The concentration of the compounds needed to observe a rescued DBA phenotype in TF1 RPS19 KD cells was determined via qRT-PCR analysis of p53 levels. The concentrations of drugs used in TF1 cells above that did not result in cell toxicity were 3.125 μΜ (Dose 3), 6.25 μΜ (Dose 2), and 25 μΜ (Dose 1). Nine different cell lines were used. The wild type (WT) TF1 and 8 TF1 KD cell lines transduced with 1 of the following plasmid: A) pSMART PGK GFP shNT-dox inducible; B) pSMART PGK GFP shRSP19_l-dox inducible; C) pSMART PGK GFP shRPS19_2-dox inducible; D) pSMART EFlalpha GFP shNT-dox inducible; I) pLKO RPS19_sh4-constitutive; J) pLKO RPS19_sh5-constitutive; K) pLKO shNT-constitutive; and L) PLKO RPS19_shl-constitutive. Cells were collected in separate tubes in a concentration of 1000

per well. Cell were centrifuged at 1,200 rpm for 5 minutes. Then, the media was aspirated and cells were resuspended in RPMI+GCMSF medium.

[0103] For the preparation of Dose 1 (final concentration on plate 25μΜ), a 50μΜ solution was made by adding 300μί of ΙΟΟμί stock to 300μί of RPMI + GCMSF + 2% DMSO (final 1% DMSO). The 50μΜ solution was further diluted to 2X when added with the cells into a well, resulting in a final concentration of 25 μΜ. For the preparation of Dose 2 (final concentration on plate 6.125 μΜ), a 12.25 μΜ solution was made by adding 75μ1. of ΙΟΟμί stock to 525μί of RPMI+GCMSF+2% DMSO. The 12.25μΜ solution was further diluted 2X when added with the cells into a well, resulting into a final concentration of 6.125μΜ. For the preparation of Dose 3 (final concentration on plate 3.125μΜ), a 6.25μΜ solution was made by adding 37.5μ1. of 100 μΐ. stock to 562.5μί of RPMI+GCMSF+2% DMSO. The 6.25μΜ solution was further diluted 2X when added with the cells into a well, resulting in a final concentration of 3.125μΜ. DMSO 1% was used for each cell type as DMSO control. [0104] A 96-well plate was used for analysis of each dose on each cell type. Fifty of each cell type and 50μί of a diluted compound are added to the corresponding wells for a total of ΙΟΟμί RPMI+GCMSF+1% DMSO, with 50,000 cells/well. The following day, 50μΙ. for RNA (25,000 cells) and 50μ1. for viability ATP assay (25,000 cells) were collected from each well.

P. Drug dilutions for administration

[0105] Prior to compound treatment, the drugs and bioactive compounds were diluted according to the parameters in Table 4. For the drugs, lOOmM stock drugs in DMSO were diluted to lOmM by mixing 0. ImM of the stock drug in DMSO with 0.9ml of DMSO to a final volume of 1.0ml. Five μΐ of the diluted drug was added to each well, and 0.5ml of media was added per well of drug. Each drug was analyzed in triplicate. Dilution to lOOOx was performed by adding 5μ1 of drug into 45μ1 of media, and the 50μ1 being added to 450μ1 of media on cells.

[0106] Table 5 provides the weight of the compound for a lOOmM dilution in 1ml of DMSO, the volume of DMSO in the lOOmM dilution, the amount of added volume, the final concentration, and the storage temperature for the dry compound.

[0107] Bioactive compounds were also used. To obtain lOOOx stock of the bioactive compounds in 1ml DMSO, 0.1 ml of 10,000X stock was combined with 0.9ml DMSO.

Bioactive compounds as diluted had the characteristics shown in Table 6. Table 6. Dilution values for bioactive com ounds

Q. Transduction of hematopoietic cells with conditional immortalization transgenes

[0108] Cryopreserved CD34+ cells (cord blood, peripheral blood, or bone marrow derived) were thawed from cryovials (typically lxl 0⁶ or 5x10⁶ cells/vial) at 37°C until completely thawed. Cells were then grown at a density of lxlO⁶ cells/mL for a period of 11 days in Hematopoietic Stem Cell expansion media DXF (Promocell) for the expansion of hematopoietic progenitor cells. Expanded CD34+ progenitor cells were then grown for an additional 3 days in SFEMII medium (Stem Cell Technologies) supplemented with the Erythroid Expansion supplement (Stem Cell Technologies). Cells can be efficiently cryopreserved at this step and are ready for viral transduction.

[0109] The Lenti-X Tet-One Inducible Expression System (Puro) (Clontech, Cat. No.

631847) was used for the transfection of 293FT cells. pLVX-TetOne-Puro vector was modified to exchange the Puro cassette with a Blasticidin (BLA), and full length cMYC, BCL-XL, BCL- XL-T2A-MYC and BCL-XL-IRES-MYC were synthesized and cloned into the pLVX-TetOne- Puro and pLVX-TetOne-BLA vectors.

[0110] 293FT cells (Clontech) were transfected with 7μg of pLVX-TetOne-Puro lentiviral vector combined with Lenti-X Packaging Single Shots (VSV-G) plasmids (Clontech). 293FT cells were transduced using the plasmid mixture. 48 hours after the transduction, the viral supernatant was collected and filtered through a 0.45μΜ PVDF unit. The viral particles were concentrated by adding 1/3 of the total volume of Lenti-X Concentrator (Clontech). The mixture was mixed well by inversion and incubated for 1-2 hours at 4°C. After precipitation, the tubes were centrifuged at 4°C for more than 45 minutes at 1,500 g to pellet the virus. Then, the supernatant was aspirated and the pellet was resuspended in PBS (to a concentration of 25-50X). Plates were coated with retronectin and incubated for more than 4 hours at RT or overnight at 4°C, using the Retronectin-transduction protocol (Clontech) according to the manufacturer's instructions. Following the Retronectin-incubation protocol (Clontech), the plates were blocked with 2% BSA/PBS for more than 30 minutes, and washed with PBS. Target cells were seeded onto the Retronectin-coated plates and the appropriate amount of virus was added. Then the plates were spun at 2,500 rpm for 1-2 hours at room temperature. Cells were incubated at 37°C for 72 hours following the infection, prior to selection.

[0111] For the selection, antibiotic was added in the cell culture medium (for example 5μg/mL of PURO). Following the death of all the non-resistant cells, the surviving cells were collected for RNA and protein assays. For RNA extraction, the cells were harvested in 500μί of TRIZOL reagent (Thermo Scientific), and incubated for 5 minutes at RT. Then, ΙΟΟμί of chloroform was added to the cells, and the mixture was vortexed and then centrifuged at 12,000 g for 15 minutes. The aqueous phase was transferred to fresh tubes and one volume of 70% ethanol was added. The TRIZOL Plus RNA purification Kit (Thermo Scientific) was used for the remaining extraction protocol according to the manufacturer's instructions. For the conversion of RNA to cDNA the High Capacity cDNA RT kit (Thermo Scientific) was used according to the manufacturer's instructions. The cDNA was analyzed with qRT-PCR to determine the fold change of knockdown cells using Taqman Fast PCR mix and Applied Biosystems probes (Thermo Scientific).

Example 2. Conditional immortalization of erythroid cells

[0112] The 2D and 3D genomic interrogation of expanding CD34+ and the differentiated erythroblasts allowed identification and understanding of the genomic compositions of genomic signaling centers in the blood. Blood can be readily obtained from diseased patients allowing target validation and compound testing on actual diseased patient cells and models thereof. Deciphering the 2D and 3D maps of normal and diseased blood provides an understanding of the disease biology and allows for screening of new and existing chemical compounds to benefit patients with unmet medical needs.

[0113] Conditional immortalization allowed for production of a large biomass, while retaining normal cell expression, size, and phenotype. Conditional immortalization was performed as described in Example 1 at day 3 of differentiation. Two samples of cord blood (CB) cells and peripheral blood (PB) CD34+ cells containing lxlO⁶ cells each were expanded for 11 days. FIG. 1 provides a summary of the timing of expansion, differentiation, and cell collection. The dash line represents DBA patient cells, and the dotted line represents the conditionally immortalized inducible DBA phenotype model. The solid line represents healthy patient cells.

[0114] Antibiotic selection was performed to select virally transformed cells. Cell collection at day 11 showed lOOxlO⁶ CB cells, and 25xl0⁶. Cell collection at day 7 of differentiation showed 400xl0⁶ CB cells, and lOOxlO⁶ peripheral blood cells. Cells conditionally immortalized at day 3 of differentiation were observed to have the capacity to provide a nearly infinite number of cord blood or peripheral blood erythroblast cells. This point in differentiation was chosen as differentiation status was synchronized at the basophilic erythroblast state and cells could be efficiently propagated and cryopreserved. Additionally, this differentiation state is relevant for studying red blood cell diseases.

A. Normal and conditionally immortalized cells possess similar globin levels

[0115] RNA-seq was used to compare the levels of expression of hemoglobin-associated genes among normal and conditionally immortalized cord blood and peripheral blood cells. Table 7 shows the Fragments Per Kilobase of transcript per Million (FPKM) for genes HBB, HBA2, HBA1, HBG1, HBG2, and HBD.

Table 7. Levels of ex ression of hemoglobin-associated enes

[0116] These results show no significant differences in the level of expression between normal and conditionally immortalized cells.

Example 3. Designing disease models

[0117] Conditionally immortalized erythroblast cells were genetically engineered via CRISPR and shRNA modulation to model DBA. This technique may be applied to model other diseases as desired. This system affords the ability to attain nearly unlimited amounts of primary cell material i.e., the biomass required for 2D and 3D genomic interrogation. As described above, DBA cell viability dramatically decreases during erythroid differentiation (days 11-18) leading to an insufficient number of DBA cells to perform 2D or 3D analysis.

[0118] To model the DBA phenotype, hematopoietic progenitor cells and erythroblasts were transduced with short hairpin RNAs (shRNAs) targeting RPS19, one of the key ribosomal proteins identified in DBA patients, and targeting p53 to rescue the DBA phenotype. shRPS19- GFPtrad (gene knock-down) was used to transduce both PB and CB conditionally immortalized erythroid cells with conditional shP53. This allowed expansion of DBA model with protection from overexpression of p53 representing an alternative model of DBA disease. GFPtrad expressing cells were magnetically enriched to avoid the need for additional cell sorting equipment. GFPtrad also allowed for rapid selection of cells when the transgene is toxic to the cells (e.g. shRPS 19). The GFP reporter was observed to be inside and outside of the cells to allow multicolor epitope flow cytometry through an anti-GFP antigen presenting cell because only the GFP channel is used. [0119] In certain embodiments, self-delivering RNA (sdRNA) is used in place of a shRNA. The hydrophobic properties of sdRNA allow it to enter cells without need of transfection reagents. Moreover, it is transient allowing more fine-tuned adjustment of knockdown through concentration of the sdRNA. sdRNA can be "washed off to return RNA to normal levels more quickly than with Dox inducible shRNAs. sdRNA may also be used both in vivo and in vitro.

[0120] Alternatively, a Cas9 protein and mRNA sgRNA CRISPR system was used to disrupt a single RPS 19 allele in both PB and CB conditionally immortalized erythroid cells with conditional shP53. The efficiency of the CRISPR (NHEJ) was assessed using the GENEART® Genomic Cleavage Detection Kit (ThermoFisher Scientific) using the standard protocol for the kit.

[0121] The DBA phenotype was then conditionally rescued during expansion and differentiation of hematopoietic DBA stem/progenitor cells toward the erythroid lineage. Two different sets of knockdowns were performed, a doxycycline inducible knockdown (mir30 based hairpin design of shRNA) and a constitutive knockdown. Once an effective shRPS19 was identified, cells were treated with compounds to determine whether the disease phenotype was rescued.

[0122] DBA cells have 2-4-fold less RPS 19 protein than healthy cells. (See, Gazda, H. T. and Sieff, C. A. (2006), British Journal of Haematology, 135: 149-157. doi: 10.1111/j.1365- 2141.2006.06268.x, which is hereby incorporated by reference in its entirety). shRNA mediated knockdown of RPS 19 was compared in conditionally immortalized cord blood erythroblast and normal cells. shRNAs were identified that resulted in 50-70% expression in the conditionally immortalized erythroblast cells, then in normal cells. Alternatively, overexpression of RPS 19 was used in DBA patient cells to rescue the DBA phenotype.

[0123] During days 1-11 of hematopoietic progenitors (CD34+) cell expansion, the number of DBA patient cells are significantly less than the number of normal patient cells as illustrated in FIG. 1. In fact, the number of DBA patient cells is about a third of the number of normal patient cells by day 11. Rescuing the DBA phenotype by inducible overexpression of RPS, RPL, or GATA1 in conditionally immortalized cells resulted in an increase in cell numbers during expansion and during differentiation. Inducible shRPS 19 allowed differentiation of DBA- positive model cells to produce the sufficient number of cells for 2D/3D analysis over the 18 days of expansion and differentiation. The conditionally immortal DBA model may be paused indefinitely to produce from 10⁹ to a nearly infinite number of cells, while the inducible DBA model produced about 10⁶ to 10⁷. Both models result in a greater number of cells than from the primary DBA cells (about 10⁴ to 10⁵). [0124] Examples below describe an immortalized cord-blood erythroblast cell line to produce a large biomass for analysis but that also retains normal cell expression, size, and phenotype.

[0125] A DBA model cell system is set up through shRNA-mediated knockdown of RPS19 or CRISPR removal of RPS 19 in the conditionally immortalized cell system and in normal cells. Table 8 summarizes the cell models of DBA that were analyzed.

Table 8. Anal zed disease models

[0126] shRNAs for RPS 19 were analyzed in normal CD34+ cells to identify shRNAs whose administration resulted in about 50% expression of RPS 19. The RPS 19 knockdown cell systems were treated with compounds to rescue the DBA phenotype, and the dose of these compounds were titrated to optimize the concentration.

A. Optimizing viral titers

[0127] HEK cell lines 293FT, 293T, and 293XT were transfected to determine which of these cell lines resulted in enhanced viral production. GFP expression was observed in all transfected cells. Viruses from each cell line were collected and tested for infectiousness. GFP expression was observed in samples of the 293XT cell line each transfected with a virus produced in one of the cell lines. GFP expression was observed in only 293XT and 293T cells and not in 293FT. In fact, virus produced in 293XT cells appeared to be the most infectious. Virus from 293FT was determined not to be infectious. Therefore 293XT cells were selected for virus production onward.

Example 4. Effects of different shRNAs on gene expression in normal CD34+ cells

[0128] shRNAs targeting RPS19, p53, or RPL11 were delivered to normal cord blood erythroblast cells using either the pLKO (constitutive) and pTRIPZ (inducible) plasmids described in Example 1. The mRNA expression of the p53 pathway proteins; progenitor factor GATA2; and differentiation factors HBB and GATA1 were measured by qRT-PCR. Cells not transduced with a shRNA as well as cells transduced with a non-targeting shRNA served as controls. A. Effects of constitutive knockdown

[0129] HBB is associated with differentiation of CD34+ cells, and p21 is associated with the p53 pathway. Table 9 shows the relative quantification of expression levels of HBB, p21, p53, RPLl 1, and RPS19 in cells transduced with plasmids pLKO-p53_sh4, pLKO_p53_sh5, pLKO_RPS19_l, pLKO_RPS19_2, pLKO_RPS19_3, pLKO_RPS19_4, and pLKO_RPS19_5 compared to levels in a control pLKO-sNT. GAPDH was used as an internal control.

Table 9. Relative uantification of ex ression levels of HBB 21 53 RPL11 and RPS19

[0130] RPS19 knockdown was observed to lead to consistent activation of the p53 pathway and decreased expression of the differentiation factors HBB and GATA1, as shown in Table 9. Cells transduced with pLKO_p53_sh4 or pLKO_p53_sh5 with an shRNA targeting p53 had a reduction in expression of p53 compared to the control. For cells transduced with

pLKO_p53_sh3, expression of other the other genes analyzed was unchanged.

[0131] To determine the effect of p53 knockdown on expression of other genes, cells were also transduced with pLKO containing p53_sh4 or p53_sh5. To determine the effect of RPS19 knockdown on expression of other genes, cells were transduced with pLKO containing RPS19_shl, RPS19_sh2, RPS19_sh3, RPS19_sh4, RPS19_sh5, and sh_NT as a control. The relative quantifications of expression of differentiation associated genes ALAS2 and HBB, immortalization factors BCLXL and MYC, and targets of interest p53 and RPS19 are shown in Table 10.

Table 10. Relative quantification of expression levels of ALAS2, HBB, BCLXL, MYC, p53, and RPS19

[0132] RPS19_sh4 and RPS19_sh5 were observed to result in a 12.8 and 5.8-fold increase in expression of p21, respectively compared to the level of expression observed for the p53 knockdowns.

B. Effects of inducible knockdown

[0133] Hematopoietic progenitor CD34+ cells were transduced with the pTRIPZ inducible lentiviral plasmids containing p53_shl, puro shNT, RPLl l l, RPL11 2, RPS19 1, or

RPS19 2. Table 11 shows the relative quantification of expression levels of HBB, p21, p53, RPL11, and RPS19 in cells transduced with inducible pTRIPZ. Table 11. Ϊ telative quantification of ex pression levels of HBB, p21 L, p53, RPL11, and RPS19

Sample Target CT RQ min max Sample Target CT RQ min max

RPS19 2 p53 25.24 1.12 0.96 1.31 RPL11 1 p53 24.70 2.31 1.91 2.80

RPS19 2 RPL11 20.55 1.46 1.09 1.96 RPL11 1 RPL11 23.75 0.23 0.15 0.35

RPS19 2 RPS19 20.73 0.77 0.66 0.89 RPL11 1 RPS19 19.66 2.30 1.76 3.00

RPS19 2 p21 26.82 1.73 1.37 2.19 RPL11 1 p21 23.66 22.15 17.89 27.43

RPS19 2 HBB 25.88 0.55 0.45 0.67 RPL11 1 HBB 26.44 0.53 0.45 0.63

RPS19 2 B-ACTIN 21.22 RPL11 1 B-ACTIN 21.73

RPS19 1 p53 25.12 1.66 1.27 2.17 puro shNT p53 25.25 1.00 0.85 1.18

RPS19 1 RPL11 20.62 1.90 1.32 2.74 puro shNT RPL11 20.94 1.00 0.80 1.25

RPS19 1 RPS19 21.89 0.47 0.41 0.54 puro shNT RPS19 20.20 1.00 0.82 1.22

RPS19 1 p21 25.51 5.87 4.31 7.98 puro shNT p21 27.46 1.00 0.74 1.36

RPS19 1 HBB 26.47 0.50 0.40 0.63 puro shNT HBB 24.87 1.00 0.78 1.28

RPS19 1 B-ACTIN 21.67 puro shNT B-ACTIN 21.07

RPL11 2 p53 25.52 2.35 1.91 2.88 p53 shl p53 26.40 0.43 0.35 0.53

RPL11 2 RPL11 23.93 0.36 0.27 0.47 p53 shl RPL11 20.90 0.98 0.84 1.14

RPL11 2 RPS19 19.86 3.58 3.22 3.99 p53 shl RPS19 20.09 1.02 0.87 1.20

RPL11 2 p21 24.62 20.43 15.77 26.46 p53 shl p21 27.42 0.98 0.80 1.20

RPL11 2 HBB 26.79 0.75 0.67 0.84 p53 shl HBB 25.25 0.73 0.59 0.91

RPL11 2 B-ACTIN 22.57 p53 shl B-ACTIN 20.99

[0134] Cells transduced with shl demonstrated lower levels of RPS19 mRNA compared to non-transduced cells. This showed that the pTRIPZ inducible shRNA was working efficiently. Expression of p21 was significantly increased in cells transduced with RPLl 1 RNA knockdown vectors pTRIPZ RPLl 1 1 and pTRIPZ RPLl 1 2, confirming the DBA-like phenotype.

[0135] The relative expression of GATA1, GATA2, MDM2, BAX, and BAG1 were also analyzed in cells transduced with a pTRIPZ dox inducible shRNA. Results are shown in Table 12. B-Actin was used as an internal control.

Table 12. Relative quantification of expression levels of GATA1, GATA2, MDM2, BAX, and BAG1

[0136] Among the p53 pathway proteins analyzed, shl transduction lead to a higher expression of BAX, P21, P53, and MDM2 mRNA, compared to sh2 transduction. Additionally, the mRNA expression levels of the progenitor protein GATA2 were higher upon shl transduction compared to sh2 transduction. mRNA levels of the differentiation associated proteins HBB and GATA1 were higher following sh2 transduction compared to their levels upon shl transduction.

C. Reversion of mortality

[0137] Early erythroblast progenitor cells derived from cord blood were transduced using lentiviral vectors harboring overexpression constructs of apoptosis-associated genes, BCLXL and MYC, under the control of a doxycycline inducible promoter. The mRNA expression of the differentiation markers ALAS2, HBB and the targets of interest, namely, p53 and RPS19 were measured using qRT-PCR, and results are shown in Table 13.

Table 13. Relative quantification of expression of RPS19, p53, HBB, ALAS2, MYC,

BCLXL, and p21

[0138] The mRNA expression of MYC and BCLXL was highest upon doxycycline induction and decreased 2, 3, 4, 5, and 6 days after doxycycline was withdrawn. The mRNA expression of the differentiation markers ALAS2 and HBB increased immediately after doxycycline was withdrawn and gradually decreased up to day 6 of doxycycline withdrawal. The mRNA expression partem of MYC and BCLXL sharply decreased 2 days after doxycycline was withdrawn as shown in Table 13. The decrease in the mRNA expression of the immortalization factors correlated with an increase in the mRNA expression of the differentiation factors. This confirms that immortality can quickly be reversed upon system shut off via retracting immortalization factors.

D. Comparing cell size after removal of Pox in cells transduced with Model 2

[0139] Cell size is shown in Table 14 during doxycycline induction, and immediately following withdrawal of doxycycline. Conditionally immortalized cells maintained in the presence of doxycycline were observed to have a cell size between 12-13.5μιη. By day 5 off doxycycline cells dramatically decreased in size to approximately 9μιη, which demonstrates a reversion of mortality and differentiation to late erythroblast stage.

Table 14. Cell size

E. Effects of knockdown of both RPS19 and shP53

[0140] Further qRT-PCR analysis of the parental cord blood IRES cell line with Dox inducible p53 (CB-IRES-p53-GFPtrad) that were separated using turbo-GFP magnetic beads is shown in Table 15. GAPDH was used as an internal control.

Table 15. Percent of knockdown

[0141] The cells administered the shRPS19 were observed to have an about 50% knockdown of RPS19, while p53 remained constant. p53 remains constant due to the NT control expressing the shRNA for p53. p53 is completely reduced in all three lines. The RPS19 shRNA constructs were normalized to the NT (shp53) control.

Example 5. Effects of different shRNAs on gene expression in conditionally immortalized cells

[0142] qRT-PCR was used to analyze expression of p53 pathway-associated genes p21, RPS19, and RPL11; differentiation marker HBB; and internal control B-ACTIN in the cord blood conditionally immortalized CD34+ (CB-CI-IRES-CD34+) cell line transduced with the pTRIPZ plasmid containing NT (Model 9), RPL1 l shl (Model 4) and RPL1 l_sh2 (Model 5) for a RPL11 knockdown, and RPS19_shl (Model 6) and RPS19_sh2 (Model 7) for a RPS19 knockdown. Results are shown in Table 16. Sample Target CT RQ min max Sample Target CT RQ min max pTRIPZ NT RPS19 20.23 1.00 0.90 1.12 RPS19 shl RPS19 22.51 0.30 0.21 0.41 pTRIPZ NT RPL11 20.98 1.00 0.88 1.13 RPS19 shl RPL 11 20.91 1.52 1.10 2.08 pTRIPZ NT p21 25.78 1.00 0.83 1.20 RPS19 shl p21 25.87 1.35 0.95 1.94 pTRIPZ NT HBB 25.78 1.00 0.79 1.27 RPS19 shl HBB 25.11 2.28 1.64 3.17 pTRIPZ NT B-ACTIN 21.59 RPS19 shl B-ACTIN 22.12

RPL11 shl RPS19 21.99 1.48 1.25 1.74 RPS19 sh2 RPS19 21.46 0.46 0.38 0.54

RPL11 shl RPL11 22.94 1.29 0.50 3.31 RPS19 sh2 RPL 11 20.78 1.23 1.07 1.42

RPL11 shl p21 25.80 4.96 3.96 6.22 RPS19 sh2 p21 26.16 0.82 0.59 1.13

RPL11 shl HBB 27.01 2.14 1.74 2.64 RPS19 sh2 HBB 25.33 1.46 1.14 1.88

RPL11 shl B-ACTIN 23.92 RPS19 sh2 B-ACTIN 21.69

RPL11 sh2 RPS19 21.45 1.28 1.09 1.50

RPL11 sh2 RPL11 22.83 0.83 0.72 0.95

RPL11 sh2 p21 25.93 2.68 2.17 3.32

RPL11 sh2 HBB 27.33 1.02 0.88 1.19

RPL11 sh2 B-ACTIN 23.17

A. Comparing gene expression levels during expansion and differentiation

[0143] qRT-PCR was used to determine the levels of expression of ALAS2, GAPDH, GATAl, GATA2, HBB, and p53 in differentiated CB-CI-IRES cells using GAPDH as an internal control as shown in Table 17.

Table 17. Re ative uantification of ex ression of GATAl, GATA2, HBB, ALAS2, and 53

[0144] Gene expression was compared for CB-CI-IRES cells during expansion and differentiation. Table 18 shows the differences in expression of differentiation markers GATAl, GATA2, and HBB and p53 in erythroid cells transduced with different lentiviruses resulting in a p53 knockdown. GAPDH was used as an internal control. Table 18. Relative uantification of ex ression levels of GATAl GATA2 HBB, and p53

[0145] Mature erythroid specific genes HBB and ALAS2 were observed to be upregulated, which confirmed that the cells were differentiated.

B. Changes in gene expression levels of GATAl and GATA2 through Day 15 of

differentiation

[0146] ChlP-seq results herein show that GATAl and GATA2 binds genomic signaling centers in both expanding and differentiated cord blood cells. Changes in expression of GATAl and GATA2 were further observed starting a day 10 (D10), then measured at days D1-D4, D7- D8, D10-D11, and D14-D15 of differentiation by qRT-PCR. Results are shown in Table 19.

Table 19. Relative uantification throu h cord blood eel differentiation

D8-Diff GATAl 1.78 1.64 1.94 GATA2 0.66 0.58 0.75

DlO-Diff GATAl 2.35 2.31 2.39 GATA2 0.34 0.32 0.37

Dl l-Diff GATAl 3.04 2.83 3.26 GATA2 0.36 0.34 0.39

D14-Diff GATAl 3.41 3.27 3.55 GATA2 0.30 0.27 0.34

D15-Diff GATAl 1.50 1.42 1.57 GATA2 0.24 0.20 0.30

[0147] GATAl was observed to increase throughout erythroid differentiation, and GATA2 was observed to decrease only at late stage erythroid differentiation.

C. Differences in expression of apoptotic genes

[0148] Gene expression of differentiation marker ALAS2 and apoptotic markers BAX and GADD45A were compared were compared between expansion Day 11 (Exp Dl 1) and Differentiation Day 7 (Diff_D7) using qRT-PCR. GAPDH was used as an internal control.

Table 20 provides the qRT-PCR results.

Table 20. Relative quantification of ALAS2, BAX, GATAl, GATA2, HBB, p21, p53, and

GADD45A

[0149] Table 20 confirms that GATAl and GATA2 are expressed in both progenitor and differentiated cells. Further, pro-apoptotic markers, such as BAX and GADD45a, were observed to be expressed at a higher level in differentiated cells compared to CD34+ progenitor cells.

D. Analyzing knockdown efficiency

[0150] The efficiency of different shRNAs were tested in a conditionally immortalized cord blood erythroblast cell system to mimic DBA. The pTRIPZ was used to transduce these cells with shl (dox inducible shRPS19) or sh2 (dox inducible shRPS19), and pLKO was used to transduce a group of cells with sh5 (shRPS19). qRT-PCR was used to determine the level of gene expression in RPS19 mRNA knockdown in each group of cells. The percent of knockdown is shown in Table 21 after 1 week (Tl), 2 weeks (T2), and 3 weeks (T3) for each shRNA. Table 21. Percent of RPS19 knockdown in CI cord blood CD34+ cells

[0151] shRNA against RPS19 possessed a 45-75% knockdown at Tl. Cells were observed to experience an increase in the percent of knockdown at T2 in each of the groups. By T3, cells became negatively selected for the knockdown.

[0152] These results show that Models 2 and 5 would be the most effective for generating the volume of biomass needed to analyze using next generation sequencing techniques, such as

ChlP-seq and RNA-seq. The estimated time to achieve enough biomass for ChIP was about one month for Model 2 and about 2.5 months for Model 5.

Example 6. Normal CD34+ progenitor cells from peripheral blood

[0153] Human CD34+ cells are derived from the peripheral blood and grown in a liquid culture. The length of time for differentiation is increased compared to the length of time for differentiation of the CD34+ cord blood cells. Positive selection of erythroblasts using magnetic beads was used to provide a more homogenous population for genome-wide studies, such as

ChlP-seq to identify genomic positions and the composition of genomic signaling centers.

A. Presence of cell surface markers in expanded peripheral blood CD34+ cells

[0154] Mapping is initially performed at differentiation stages most affected in the DBA phenotype to build a reference gene expression pattern for human erythropoiesis, such as the BFU/ CFU and early erythroblast stages. By day 7 of expansion, BFU/CFU cell markers were observed to be present, and by day 7 of differentiation, early erythroblast cell markers were observed to be present. BFU and CFU markers were present in the expanded CD34+ cells described above. BFU cells are known to express CD34 and CD71, and CFU cells are known to express high levels of CD36 and CD71. qRT-PCR was used to measure expression throughout the expansion of the CD34+ peripheral blood cells. Relative quantification of the level of expression of CD71 transferring receptor from day 4 of expansion (exp-D4) to day 1 of differentiation (Diff-Dl) is shown in Table 22.

[0155] Relative quantification of the level of expression of CD36 from day 4 of expansion (exp-D4) to day 1 of differentiation (Diff-Dl) are shown in Table 23. Table 23. Relative quantification (RQ) of expression of CD36

[0156] Therefore, the expanded cells expressed markers known to be present in BFU and CFU cells and are the most effected by the DBA phenotype.

Example 7. Determining the time point for Glycophorin A positive (GYPA+) selection

[0157] qRT-PCR was used to show changes in the levels of expression of GATAl and GATA2 in both peripheral and cord blood cells to determine a time point for GYPA+ selection. GAPDH was used as an internal control. Table 24 provides the relative quantification of the expression of these genes.

Table 24. Relative quantification (RQ) of GATAl and GATA2

[0158] While GATAl levels gradually increased during erythroid differentiation in both cord blood and peripheral blood cells, GATA2 levels sharply decreased 60-70% during erythroid differentiation in peripheral cells. Cord blood cells only showed sharp decrease in GATA2 at day 10 of differentiation, while a sharp increase was observed in the peripheral blood cells on day 2 of differentiation.

A. Enriching erythroid cells expressing GATAl

[0159] GYPA+ selection was used at day 12 to enrich peripheral erythroid cells expressing GATAl. GYPA+ selection was observed to increase the ratio of GATAl to GATA2 over 3-fold, i.e. 40X more GATAl at day 12, as shown in Table 25. Table 25. Relative quantification (RQ) of GATAl and GATA2 after GYPA+ selection

[0160] As shown above, GATA2 levels sharply decreases at day 1 (Dl) of differentiation. GATAl levels were observed to gradually increase during erythroid differentiation. Day 7 (D7) of differentiation was chosen as a time point for GYPA+ selection because it provided an early erythroid population, i.e., GATA2 levels were observed to be low. Seven days of differentiation also allowed for a sufficient increase in biomass for ChIP analysis. The erythroid cells were enriched at day 12 (D 12) to determine the effectiveness late in differentiation.

Example 8. Modeling disease using self-delivering RNA (sdRNA)

[0161] Dox removal initiates normal differentiation, and a window for recovery of levels of p53 after efficient knockdown (>95%) with a shP53 should be 3-7 days. As an alternative to using shRNAs, a DBA model was designed using self-delivering RNA (sdRNA) to knockdown p53. sdRNAs provide a transient p53 knockdown, and the level of knockdown can be more easily optimized. sdRNAs targeting additional disease targets, such as RPS 19, are also designed and tested.

[0162] To validate the efficacy of sdRNA to temporarily alleviate cell death due to RPS19 knockdown, p53 knockdown efficiency and reversibility were assessed.

[0163] sdRNAs targeting p53 were introduced to CB-IRES cells. Gene expression of p53 was measured using qRT-PCR at either 48 or 72 hours. Results for the sdRNAs at different concentrations is shown in Table 26.

Table 26. Relative quantification (RQ) of p53 during titration of p53 sdRNAs

[0164] For about 50% knockdown, the sdRNA administered at 2μΜ for 72 hours was chosen for further analysis. [0165] To determine reversibility of the knockdown, CB-IRES cells were treated with sdRNAs for 48 hours, then washed with PBS. p53 levels were monitored 5 days after initial wash. Full recovery knockdown was observed for both 4μΜ and 2μΜ of sdRNA.

Example 9. Determining genomic position and composition of genomic signaling centers in erythroid cells

[0166] Changes in the composition and changes in genomic signaling centers with and without compound treatment were observed in engineered disease models and patient disease cells.

A. ChlP-seq targets

[0167] Table 27 provides the list of ChlP-seq targets analyzed herein.

Table 27. ChlP-se tar ets for human erythroblasts

B. Changes in genomic signaling centers through cell differentiation

[0168] To produce 3D gene regulatory maps, the following steps are followed: optimization of tissue culture conditions for growth and expansion of CD34+ cells, identification of genomic positions and protein composition of genomic signaling centers in normal CD34+ cells in absence and presence of compounds, determination of genome architecture in these cells, and identification of genomic positions and protein composition of genomic signaling centers in a human model of a disease (e.g., erythroblast cells with DBA from shRNA knockdown of RPS19) or primary disease patient samples in the presence and absence of compounds, and demonstration of the capability of compounds to modulate expression of gene targets implicated in the development of a disease by leveraging genomic signaling centers. The map of the progenitor culture may also be used to predict whether a response would occur or not when a drug is administered.

[0169] To determine the protein composition and genomic position of genomic signaling centers in normal primary CD34+ and erythroblasts cells, ChlP-seq and RNA-seq were performed and results are shown for the insulated neighborhoods containing GADD45A, RPS19, and CDK 1A. The presence of H3K27Ac, GATA1, GATA2, p53, and serl5 (p53 phosphorylated at serine 15) within these insulated neighborhoods is shown. Cells were collected at day 11 of expansion prior to erythroid differentiation and day 18 after 7 days of differentiation.

[0170] The binding profile in FIG. 2A revealed a significant increase in p53 accumulation at the GADD45A promotor during differentiation. This appears consistent to what has been observed at other apoptotic genes (e.g. BAX). p53 was also abundant at the RPS19 promotor, as shown in FIG. 2B, but levels were similar pre- and post-differentiation. In FIG. 2C, a decrease in the presence of GATA1 was observed, while a decrease in expression of RPS 19 was also observed from RNA-seq results.

[0171] Therefore, FIGs. 2A-2C provided evidence that GATA1/GATA2 is co-expressed at both El 1 and D7 and p53 binds the RPS19 and GADD45a promoter. Since these data are from wildtype cells, results indicate a mechanism by which apoptotic genes are 'poised' to be triggered in this lineage. Since present results do not show a clear timepoint for the switch from GATA2 to GATA1, further differentiation may be needed.

C. Binding profiles

[0172] Collating the results of FIGs. 2A-2C showed that p53 was bound to more peaks in erythroid cells compared to hematopoietic cells (1485 vs. 743). A majority of p53 was also observed to bind non-promoter regions in hematopoietic stem cells. In fact, 3.6% of the p53 peaks are in exon regions, 23.6% are in promoter regions, 29.9% are in intron regions, and 42.9% are in intergenic regions.

[0173] For all the p53 binding sites, there is significant overlap between p53 binding in hematopoietic stem cells and erythroblasts. Five hundred seventy-eight of the p53 peaks were shared in both hematopoietic stem cells and erythroblasts. 3.6% of the p53 peaks are in exon regions, 20.8% are in promoter regions, 31.5% are in intron regions, and 44.1% are in intergenic regions. One hundred one of the 385 binding sites found in promoter regions were shared by hematopoietic stem cells and erythroblasts.

D. Gene ontology (GO) analysis of the p53 binding sites

[0174] FIGs. 3A-3D show the loglO p-values for genomic annotation of the binding sites of p53 (loglO p-value < 4 were considered significant). FIG. 3 A shows that most of the binding sites in a hematopoietic stem cell promoter region are likely associated with cellular response to DNA damage stimulus and response to UV exposure. Alternatively, these binding sites may be associated with signal transduction by p53 class mediator, signal transduction in response to DNA damage, mitotic DNA damage checkpoint, mitotic DNA integrity checkpoint, apoptotic signaling pathway, intrinsic apoptotic signaling pathway, cellular response to external stimuli, and mitotic cell cycle checkpoint in descending frequency.

[0175] FIG. 3B shows that the p53 binding sites of erythroid cells are most likely located in promoter regions associated with cellular response to DNA damage stimulus. Alternatively, the binding sites are associated with cellular response to DNA damage stimulus, DNA damage response and signal transduction by p53, class mediator resulting in cell cycle arrest, signal transduction by p53 class mediator, signal transduction involved in DNA damage checkpoint, mitotic Gl DNA damage checkpoint, mitotic DNA damage checkpoint, mitotic DNA integrity checkpoint, signal transduction involved in cell cycle checkpoint, and signal transduction in response to DNA damage in descending frequency.

[0176] In contrast to p53 binding sites in promoter regions of both hematopoietic stem cells and erythroid cells, FIG. 3C shows that p53 binding sites in non-promoter regions of hematopoietic stem cells were observed in regions of the genome most likely associated with the purine deoxyribonucleoside triphosphate catabolic process. Alternatively, p53 binding sites were also in non-promoter regions of hematopoietic stem cells most likely associated with divalent inorganic cation homeostasis, calcium ion homeostasis, cellular divalent inorganic cation homeostasis, regulation of homeostatic process, cellular cation homeostasis, release of sequestered calcium ion into cytosol, regulation of sequestering calcium ion, cellular calcium ion homeostasis, and dendritic cell differentiation in descending frequency.

[0177] Further, FIG. 3D shows that p53 binding sites in non-promoter regions of erythroid cells were most frequent in regions most likely associated with regulation of phopholipase C activity. P53 binding sites in non-promoter regions of erythroid cells were also potentially associated with regulation of phospholipase activity, cytosolic calcium ion homeostasis, positive regulation of phospholipase C activity, positive regulation of phospholipase activity, positive regulation of lipase activity, regulation of DNA biosynthesis process, regulation of transcription from RNA polymerase II, promoter in response to stress, and cellular response to decrease oxygen levels in descending frequency.

[0178] Therefore, p53 was observed to bind the promoter regions of apoptotic genes in both hematopoietic stem cells and differentiated erythroid cells, and calcium ion homeostasis-related genes are enriched in non-promoter p53 peaks.

E. Overlap with active enhancers

[0179] Each identified p53 non-promoter binding site was extended 500 base pairs upstream and downstream, and the percentage of overlap of these regions with H3K27Ac, a histone modification associated with active enhancers, was determined. Of the 566 p53 binding sites observed in non-promoter regions of hematopoietic stem cells, 47.5% of them overlapped with H3K27Ac. Of the 1170 p53 binding sites observed in non-promoter regions of erythroid cells, 37.7% overlap with H3K27Ac. Further no enriched biological processes were observed through quantification of genomic annotation of the p53 binding sites overlapping with H3K27Ac.

Example 10. Compound treatment of conditionally immortalized and normal CD34+ cells

[0180] The conditionally immortalized CD34+ cord blood primary cells (CB-CI-IRES- CD34+) and TF1 cells were treated with compounds to rescue the phenotype of DBA in the disease model. Doses of the compounds were titrated to determine the concentration needed to stimulate a response in the cells.

A. Cell toxicity

[0181] A cell viability ATP assay was performed with different concentrations of Nutlin 3, Nutlin 3a, pifithrin alpha (pifithrin-a), and cyclic pifithrin-a to determine the IC50. Cells treated with DMSO served as a control. Amount of ATP for concentrations of 100, 50, 25, 12.5, 6.252, 3.125, 1.5, 0.7, 0.35, 0.175, and 0.08625 μΜ are shown in Table 28. Measurements were performed in triplicate.

Table 28. ATP cell viability assay for CD34+ cells

[0182] A cell viability assay was also carried out for TF1 cells to determine the optimal concentration prior to cell toxicity for each compound. Results are shown in Table 29.

Table 29. ATP cell viability assay for TF1 cells

1 1292289 1399978 1333723 1374544 1386589 1380129 1323936 1380181 1379437 1328647 1266411

2 1394301 1395554 1425739 1457698 1399051 1358414 1425034 1423050 1425399 1457633 1417713

3 1313065 1345794 1382178 1366440 1353389 1354916 1412128 1401192 1367940 1400082 1369050

Avg. 1333218 1380442 1380547 1399561 1379676 1364486 1387033 1401474 1390925 1395454 1351058

Nutlin 3a

1 23934 1246588 1518028 1340587 1460412 1326585 1501481 1465593 1501089 1387750 1477730

2 23542 1204019 1486252 1403123 1231790 1450508 1426000 1321104 1461456 1481279 1426195

3 27797 1331152 1420453 1389877 1364965 1385714 1398203 1325541 1375796 1406790 1238001

Avg. 25091 1260586 1474911 1377862 1352389 1387602 1441895 1370746 1446114 1425273 1380642

Nutlin 3

27666 879100 1441451 1455179 1468908 1409570 1412845 1421014 1391417 1405054 1354903

29245 933388 1512469 1541675 1560088 1515470 1497031 1519190 1520129 1512482 1472614

23602 703030 1435866 1457750 1481632 1443604 1447193 1423442 1329299 1422972 1411853

Avg. 26837 838506 1463262 1484868 1503543 1456215 1452356 1454549 1413615 1446836 1413123

B. Dose titration

[0183] Dose titration was performed as described in Example 1. Three doses of each compound that did not result in cell toxicity in the results above were chosen to determine the amount of the compound needed to elicit a response in RPS19 knockdown cells that resulted in a rescued DBA phenotype by reducing elevated p53 levels caused by the DBA phenotype.

Phenotype rescue was determined by measuring p53 levels by qRT-PCR analysis. It was expected that the low dose (3.125 μΜ) would result in no change in p53, that the medium dose (6.25 μΜ) would result in some change in p53, and the maximal dose would result in a decrease in p53. Wildtype TFl or conditionally immortalized erythroblast cord blood cells were transduced with either pSMART PGK GFP shNT (dox inducible), pSMART PGK GFP shRPS19_l (dox inducible), pSMART PGK GFP shRPS 19_2 (dox inducible), pSMART EF1 alpha GFP shNT (dox inducible), pLKO RPS 19_sh4 (constitutive), pLKO RPS 19_sh5 (constitutive), pLKO shNT (constitutive), or pLKO RPS19_shl (constitutive). An aliquot of knockdown cells treated only with DMSO was used as a control. The 8 samples of knockdown cells are also compared to a wildtype TFl cell control. Knockdown resulted in the cell densities shown in Table 30.

C. Changes in biological activity

[0184] Compound dilutions were performed as described in Example 1. Compound Nutlin 3, Pifithrin-a, Nutlin 3a, and Cyclic pifithrin-a were administered in a 10 μΜ dose. In a 6-well format, there were 1 million cells/well = 1 million cells/ml = 1000 cells/μΐ. Eight ml of

RPMI+GMCSF was added to each well. Table 31 shows changes in cell viability. Table 31. Fold chan es in cell viabilit

[0185] TFl cells did not significantly respond to p53 pathway activators Nutlin 3/ Nutlin 3a or inhibitors Pifithrin-alpha/cyclic. CD34+ cells treated with Nutlin 3 and Nutlin 3a had reduced biological activity, while Pifithrin had little or no effect.

IV. EQUIVALENTS AND SCOPE

[0186] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments in accordance with the invention described herein. The scope of the present invention is not intended to be limited to the above Description, but rather is as set forth in the appended claims.

[0187] In the claims, articles such as "a," "an," and "the" may mean one or more than one unless indicated to the contrary or otherwise evident from the context. Claims or descriptions that include "or" between one or more members of a group are considered satisfied if one, more than one, or all of the group members are present in, employed in, or otherwise relevant to a given product or process unless indicated to the contrary or otherwise evident from the context. The invention includes embodiments in which exactly one member of the group is present in, employed in, or otherwise relevant to a given product or process. The invention includes embodiments in which more than one, or the entire group members are present in, employed in, or otherwise relevant to a given product or process.

[0188] It is also noted that the term "comprising" is intended to be open and permits but does not require the inclusion of additional elements or steps. When the term "comprising" is used herein, the term "consisting of is thus also encompassed and disclosed.

[0189] Where ranges are given, endpoints are included. Furthermore, it is to be understood that unless otherwise indicated or otherwise evident from the context and understanding of one of ordinary skill in the art, values that are expressed as ranges can assume any specific value or subrange within the stated ranges in different embodiments of the invention, to the tenth of the unit of the lower limit of the range, unless the context clearly dictates otherwise.

[0190] In addition, it is to be understood that any particular embodiment of the present invention that falls within the prior art may be explicitly excluded from any one or more of the claims. Since such embodiments are deemed to be known to one of ordinary skill in the art, they may be excluded even if the exclusion is not set forth explicitly herein. Any particular embodiment of the compositions of the invention (e.g., any antibiotic, therapeutic or active ingredient; any method of production; any method of use; etc.) can be excluded from any one or more claims, for any reason, whether or not related to the existence of prior art. [0191] It is to be understood that the words which have been used are words of description rather than limitation, and that changes may be made within the purview of the appended claims without departing from the true scope and spirit of the invention in its broader aspects.

[0192] While the present invention has been described at some length and with some particularity with respect to the several described embodiments, it is not intended that it should be limited to any such particulars or embodiments or any particular embodiment, but it is to be construed with references to the appended claims so as to provide the broadest possible interpretation of such claims in view of the prior art and, therefore, to effectively encompass the intended scope of the invention.

Claims

1. A conditionally immortal cell system for simulating at least one disease state, the cell system comprising:

(a) a population of CD34+ cells expressing at least one anti-apoptotic gene controlled by an inducible promoter; and

(b) at least one exogenous biomolecule to modulate expression of at least one disease- related gene selected from the group consisting of: RPS19, RPS14, RPL11, RPS24, RPS17, RPL35A, RPL5, RPS7, RPS 10, RPS26, RPL26, RPL15, RPS29, TSR2, RPS28, RPL27, RPS27, GATA1, and a combination thereof.

2. The cell system of claim 1, wherein the at least one anti-apoptotic gene comprises: B-cell lymphoma extra-large (BCL-XL), c-Myc (MYC), or a combination thereof.

3. The cell system of claim 1 or claim 2, further comprising at least one exogenous biomolecule to modulate the activity or expression of p53.

4. The cell system of any one of claims 1-3, wherein the at least one exogenous biomolecule comprises a short-hairpin RNA (shRNA), a self-delivering RNA (sdRNA), or a combination thereof.

5. The cell system of any one of claims 1-3, wherein the at least one exogenous biomolecule comprises a CRISPR system.

6. The cell system of claim 4 or claim 5, wherein the exogenous biomolecule is in the form of a vector or plasmid which facilitates the formation of a final exogenous biomolecule in the target cell system which modulates the expression of the at least one disease-related gene.

7. The cell system of claim 6, wherein the vector is a lentiviral vector.

8. The cell system of any one of claims 1-7, wherein the activity of the exogenous biomolecule is controlled by an inducible promoter.

9. The cell system of any one of claims 1-7, wherein the exogenous biomolecule is constitutively active.

10. The cell system of any one of claims 1-9, wherein the population of CD34+ cells is from a healthy subject.

11. The cell system of any one of claims 1-9, wherein the population of CD34+ cells is from a diseased subject.

12. The cell system of any one of claims 1-11, wherein the population of CD34+ cells is hematopoietic stem cells.

13. The cell system of any one of claims 1-12, wherein the population of CD34+ cells

originates from a source selected from the group consisting of: umbilical cord blood, peripheral blood frozen stock, fresh peripheral blood, and combinations thereof.

14. The cell system of any one of claims 1-13, wherein the disease state is Diamond- Blackfan anemia (DBA).

15. The cell system of any one of claims 1-13, wherein the disease state is 5q- myelodysplasia.

16. A method of preparing a conditionally immortal cell system for simulating at least one disease state, the method comprising:

(a) introducing into a population of CD34+ cells at least one anti-apoptotic gene,

wherein the at least one anti-apoptotic gene is under the control of a promoter that is inducible by contacting the cells with a factor, and wherein removing the factor from the cells reduces expression of the at least one anti-apoptotic gene; and

(b) introducing into the population of CD34+ cells at least one exogenous biomolecule to modulate the level of expression of at least one disease-related gene selected from the group consisting of: RPS19, RPS14, RPL11, RPS24, RPS17, RPL35A, RPL5, RPS7, RPS10, RPS26, RPL26, RPL15, RPS29, TSR2, RPS28, RPL27, RPS27, GATA1, and a combination thereof, thereby simulating the disease state in the conditionally immortal cell system.

17. The method of claim 16, wherein the at least one anti-apoptotic gene comprises B-cell lymphoma extra-large (BCL-XL), c-Myc (MYC), or a combination thereof.

18. The method of claim 16 or claim 17, wherein the at least one anti-apoptotic gene is delivered using a lentiviral vector.

19. The method of any one of claims 16-18, wherein the factor is an antibiotic.

20. The method of claim 19, wherein the antibiotic is doxycycline.

21. The method of any one of claims 16-20, further comprising introducing into the population of CD34+ cells at least one exogenous biomolecule to modulate the activity or expression of p53.

22. The method of any one of claims 16-21, wherein the at least one exogenous biomolecule comprises a short-hairpin RNA (shRNA), a self-delivering RNA (sdRNA), or a combination thereof.

23. The method of any one of claims 16-21, wherein the at least one exogenous biomolecule comprises a CRISPR system.

24. The method of claim 22 or claim 23, wherein the exogenous biomolecule is in the form of a vector or plasmid which facilitates the formation of a final exogenous biomolecule in the target cell system which modulates the expression of the at least one disease-related gene.

25. The method of claim 22 or claim 23, wherein the activity of the exogenous biomolecule is controlled by an inducible promoter.

26. The method of claim 22 or claim 23, wherein the exogenous biomolecule is constitutively active.

27. The method of any one of claims 16-26, further comprising removing the factor from the population of CD34+ cells, thereby differentiating the population of CD34+ cells.

28. The method of claim 27, wherein differentiating the population of CD34+ cells results in a greater number of erythroid cells than the population of CD34+ cells without introducing the at least one anti-apoptotic gene.

29. The method of any one of claims 16-28, wherein the population of CD34+ cells originates from a source selected from the group consisting of: umbilical cord blood, peripheral blood frozen stock, fresh peripheral blood, and combinations thereof.

30. The method of any one of claims 16-29, wherein the disease state is Diamond-Blackfan anemia.

31. The method of any one of claims 16-29, wherein the disease state is 5q- myelodysplasia.

32. A method comprising:

(a) providing a cell system of any one of claims 1 -15;

(b) contacting the cell system with a compound; and

(c) characterizing the effect of the compound on the cell system.

33. The method of claim 32, wherein the characterizing step comprises characterizing a level of altered gene expression of a disease-related gene in the cell system resulting from contacting the cell system with the compound.

34. The method of claim 33, wherein the characterizing step further comprises comparing the altered gene expression of the disease-related gene in the cell system with a corresponding level of gene expression in a cell system of a normal state.

35. The method of claim 33 or claim 34, wherein the altered gene expression is characterized using RNA-seq.

36. The method of any one of claims 32-35, wherein the characterizing step comprises

characterizing an altered binding profile of a genomic signaling center in the cell system resulting from contacting the cell system with the compound; wherein the genomic signaling center is in at least one insulated neighborhood comprising a disease-related gene.

37. The method of claim 36, wherein the characterizing step further comprises comparing the altered binding profile of the genomic signaling center in the cell system with a corresponding GSC binding profile in a cell system of a normal state.

38. The method of claim 36 or claim 37, wherein the altered binding profile is characterized using ChlP-seq.

39. The method of any one of claims 33-38, wherein the at least one disease-related gene is selected from the group consisting of RPS19, RPS14, RPLl l, RPS24, RPS17, RPL35A, RPL5, RPS7, RPS 10, RPS26, RPL26, RPL15, RPS29, TSR2, RPS28, RPL27, RPS27, GATA1, and a combination thereof.

40. The method of any one of claims 33-38, wherein the at least one disease-related gene is selected from the group consisting of P53, BAX, P21, GADD45A, CDK 1A, BAG1, MDM2, and a combination thereof.

41. The method of any one of claims 32-40, wherein the compound is selected from the group consisting of nutlin 3, pifithrin-a (PFT-a), nutlin 3a, cyclic pifithrin-a, and a combination thereof.

42. The method of any one of claims 32-41, further comprising a step of calculating an optimal concentration of the compound.

43. The method of any one of claims 32-42, further comprising a step of calculating a half maximal effective concentration (EC50) of the compound.

44. The method of any one of claims 32-43, further comprising repeating at least one of the providing, contacting or characterizing steps to screen a plurality of compounds.