**4.** *De novo* **generation of HSC**

Considering the interest in HSC expansion for treatment of both malignant and nonmalignant diseases as well as their use in gene therapy and the difficulty to obtain *ex vivo* expansion of HSC without loss of their regeneration capacities, relevant methods to produce *de novo* HSC have emerged.

#### **4.1 Obtaining HSC from ESC**

One of these methods was initiated 20 years ago when ESC could be cultivated *in vitro* and directed to generate hematopoietic cells (Wiles and Keller, 1991). Since then, culture conditions were constantly optimized and allowed the differentiation into specific hematopoietic lineages such as erythroid and myeloid lineages, T and B lymphocytes and megakaryocytes (for review see Sakamoto et al., 2010). These protocols were then adapted to human (h) ESC. These cells like their murine counterparts, are karyotypically stable, capable of prolonged self-renewal, and might differentiate into most cell types. These properties might be exploited for therapeutic benefits to cure many human degenerative diseases and resulted in intense biomedical studies.

Different methods were established to generate hematopoietic progenitors and specific lineages from mouse ESC including embryoid bodies formation, coculture with stromal cells, and direct differentiation in coated plates using a mixture of cytokines and growth factors without stromal cells (Tian and Kaufman, 2008). These protocols were then optimized for efficient differentiation of hESC into early mesodermal cells (Bernardo et al., 2011) and for obtaining defined hematopoietic precursors from ES cells (Chiang and Wong, 2011; Salvagiotto et al., 2011).

The ultimate goal of these strategies is to produce HSC capable of robust, long-term, multilineage engraftment to alleviate blood cells diseases; however the numbers and the capacities of the *de novo* cells generated are not quite sufficient to fulfill the clinical challenge. At present, multipotent hematopoietic progenitors (short-term HSC) with limited engrafting ability in transplanted mice were obtained (Woods et al., 2011). Other groups reported efficient generation of cells that mostly produce the myeloid lineage following long term engraftment or produce CD34+ hematopoietic precursors that have phenotype similar

Searching for the Key to Expand Hematopoietic Stem Cells 227

SOX2, NANOG, and LIN28 (Yu et al., 2007). Krüppel-like transcription factors (Klf2 and Klf5) and the orphan nuclear receptor, Esrrb, can replace Klf4 (Nakagawa et al., 2008 and for

The derived iPSC exhibited typical ESC morphology and were similar to ESC in their regenerative potential (Takahashi and Yamanaka, 2006) and their capacity to differentiate into cells of all three germ layers, the ectoderm, mesoderm, and endoderm. Because iPSC are generated without the need to destroy an embryo, their discovery has further energized the field of regenerative medicine and stem cell biology. Patient-specific therapeutic cells derived from induced pluripotent stem iPSC may bypass the ethical issues associated with ESC and avoid potential immunological reactions associated with allogenic transplantation. These human disease–specific iPSC provide a unique and previously unavailable resource

The therapeutical hope of iPSC is based on three issues: 1) The ability to generate iPSC from any tissue of the organism, and further differentiate them according to the patient needs, particularly into a wide range of primary human cell types, many of which are unavailable for routine use; 2) The ability to generate iPSC from patients with any disease; 3) The

The therapeutic potential of such iPSC (schematized in Fig. 5) was demonstrated in a proofof-principle study using a humanized sickle cell anemia mouse model (Hanna et al., 2007). In this study, mice could be rescued after transplantation with hematopoietic progenitors obtained *in vitro* from autologous iPSC. This was achieved after correction of the human

Also in mice, Xu et al. cured hemophilia by transplantation of cells that were generated from murine wild-type iPSC. These murine experiments suggest that human iPSC can be utilized for regenerative and therapeutic applications (Xu et al., 2009). Most recently, patientspecific iPSC have been established. Raya et al. reprogrammed dermal fibroblasts and/or epidermal keratinocytes of Fanconi anemia patients to generate iPSC, which were genetically corrected with lentiviral vectors encoding FANCA or FANCD2, to obtain hematopoietic progenitors of the myeloid and erythroid lineages that are phenotypically normal, that is, disease-free (Raya et al., 2009). Similar strategies were performed to correct the Hurler syndrome (Tolar et al., 2011) and for the production of macrophages from iPSCs

for studying the pathophysiology of various important human diseases.

possibility of using patient-derived iPSC for drug development.

Fig. 5. Studies performed to validate the therapeutic potential of iPSC.

which were resistant to HIV infection (Kambal et al., 2011).

sickle hemoglobin allele by gene-specific targeting.

review see Feng et al., 2009).

to adult HSC but might best correspond to the embryonic stage of yolk-sac, aortogonadalmesonephros (AGM), and/or fetal liver stage of hematopoiesis (Melichar et al., 2011; Narayan et al., 2006 and for review : Tian and Kaufman, 2008). More recently, the polycomb group protein Bmi1 was shown to promote more than 100-fold increase of hematopoietic cell development from ESC (Ding et al., 2011).

Since short-term HSC could be generated from ESC, an attractive option to increase the number of clinically competent HSC would be to find a molecule that dedifferentiate from short-term or mature hematopoietic cells to the long-term HSC population. Such a strategy might be valuable, since de-differentiation of somatic cells mediated by a chemical has been achieved in other systems. This is the case for reversine or 2-(4-morpholinoanilino)-6-cyclohexylaminopurine. This chemical compound was reported to induce reversal of mouse myoblast cell line, C2C12, to become multipotent progenitor cells, which can re-differentiate into osteoblasts and adipocytes (Chen et al., 2004). The de-differentiation activity of reversine however is not conserved across all cell lineages, since in certain cell types, it acts as a potent differentiation-inducing molecule (D'Alise et al., 2008).

To support the generation of long-term repopulating HSC from mouse ESC, other groups tested intrinsic regulators of adult HSC (Schuringa et al., 2004; Wang et al., 2005b). However, the use of many of these compounds, such as HoxB4, did not improve the expected engraftment efficiency *in vivo* (Wang et al., 2005a).

An unfavorable complication for the use of ESC in producing HSC is that lifelong use of drugs is required to prevent rejection of the transplanted cells. In order to make ESC practical for therapeutic use, it would be necessary to create a new stem cell line for each patient that needs treatment. Serious technical and ethical problems are associated with this issue.

### **4.2 Obtaining HSC from induced pluripotent stem cells**

An alternative to the utilization of ES cells to produce *de novo* HSC arise from one of the most transformative accomplishments performed in the last years: the discovery that transient overexpression of a small number of defined transcription factors can reprogram the differentiated cells and become pluripotent populations. These cells are commonly referred to as Induced Pluripotent Stem Cells (iPSC) and have definitively broken the dogma commonly accepted that differentiated cell types generally lack the ability to revert back to a less specialized state.

#### **4.2.1 Reprogramming somatic cells to pluripotency**

The direct reprogramming of somatic cells to pluripotency was demonstrated in 2006, when Takahashi and Yamanaka converted adult mouse fibroblasts to iPSC by overexpressing four transcription factors: octamer-binding transcription factor 4 (OCT4), sex determining region Y-box 2 (SOX2), Kruppel-like factor 4 (KLF4), and cytoplasmic Myc (c-MYC) in mouse embryonic fibroblasts using retroviruses (Takahashi and Yamanaka, 2006). The transcription factors originally used for reprogramming differentiated cells are not stringently necessary to achieve this process as some of them can be replaced by other factors. Yu et al*.* were able to reprogram human fibroblasts with a distinct set of transcription factors comprising OCT4,

to adult HSC but might best correspond to the embryonic stage of yolk-sac, aortogonadalmesonephros (AGM), and/or fetal liver stage of hematopoiesis (Melichar et al., 2011; Narayan et al., 2006 and for review : Tian and Kaufman, 2008). More recently, the polycomb group protein Bmi1 was shown to promote more than 100-fold increase of hematopoietic

Since short-term HSC could be generated from ESC, an attractive option to increase the number of clinically competent HSC would be to find a molecule that dedifferentiate from short-term or mature hematopoietic cells to the long-term HSC population. Such a strategy might be valuable, since de-differentiation of somatic cells mediated by a chemical has been achieved in other systems. This is the case for reversine or 2-(4-morpholinoanilino)-6-cyclohexylaminopurine. This chemical compound was reported to induce reversal of mouse myoblast cell line, C2C12, to become multipotent progenitor cells, which can re-differentiate into osteoblasts and adipocytes (Chen et al., 2004). The de-differentiation activity of reversine however is not conserved across all cell lineages, since in certain cell types, it acts

To support the generation of long-term repopulating HSC from mouse ESC, other groups tested intrinsic regulators of adult HSC (Schuringa et al., 2004; Wang et al., 2005b). However, the use of many of these compounds, such as HoxB4, did not improve the

An unfavorable complication for the use of ESC in producing HSC is that lifelong use of drugs is required to prevent rejection of the transplanted cells. In order to make ESC practical for therapeutic use, it would be necessary to create a new stem cell line for each patient that needs treatment. Serious technical and ethical problems are associated with this

An alternative to the utilization of ES cells to produce *de novo* HSC arise from one of the most transformative accomplishments performed in the last years: the discovery that transient overexpression of a small number of defined transcription factors can reprogram the differentiated cells and become pluripotent populations. These cells are commonly referred to as Induced Pluripotent Stem Cells (iPSC) and have definitively broken the dogma commonly accepted that differentiated cell types generally lack the ability to revert

The direct reprogramming of somatic cells to pluripotency was demonstrated in 2006, when Takahashi and Yamanaka converted adult mouse fibroblasts to iPSC by overexpressing four transcription factors: octamer-binding transcription factor 4 (OCT4), sex determining region Y-box 2 (SOX2), Kruppel-like factor 4 (KLF4), and cytoplasmic Myc (c-MYC) in mouse embryonic fibroblasts using retroviruses (Takahashi and Yamanaka, 2006). The transcription factors originally used for reprogramming differentiated cells are not stringently necessary to achieve this process as some of them can be replaced by other factors. Yu et al*.* were able to reprogram human fibroblasts with a distinct set of transcription factors comprising OCT4,

cell development from ESC (Ding et al., 2011).

as a potent differentiation-inducing molecule (D'Alise et al., 2008).

expected engraftment efficiency *in vivo* (Wang et al., 2005a).

**4.2 Obtaining HSC from induced pluripotent stem cells** 

**4.2.1 Reprogramming somatic cells to pluripotency** 

issue.

back to a less specialized state.

SOX2, NANOG, and LIN28 (Yu et al., 2007). Krüppel-like transcription factors (Klf2 and Klf5) and the orphan nuclear receptor, Esrrb, can replace Klf4 (Nakagawa et al., 2008 and for review see Feng et al., 2009).

The derived iPSC exhibited typical ESC morphology and were similar to ESC in their regenerative potential (Takahashi and Yamanaka, 2006) and their capacity to differentiate into cells of all three germ layers, the ectoderm, mesoderm, and endoderm. Because iPSC are generated without the need to destroy an embryo, their discovery has further energized the field of regenerative medicine and stem cell biology. Patient-specific therapeutic cells derived from induced pluripotent stem iPSC may bypass the ethical issues associated with ESC and avoid potential immunological reactions associated with allogenic transplantation. These human disease–specific iPSC provide a unique and previously unavailable resource for studying the pathophysiology of various important human diseases.

The therapeutical hope of iPSC is based on three issues: 1) The ability to generate iPSC from any tissue of the organism, and further differentiate them according to the patient needs, particularly into a wide range of primary human cell types, many of which are unavailable for routine use; 2) The ability to generate iPSC from patients with any disease; 3) The possibility of using patient-derived iPSC for drug development.

The therapeutic potential of such iPSC (schematized in Fig. 5) was demonstrated in a proofof-principle study using a humanized sickle cell anemia mouse model (Hanna et al., 2007). In this study, mice could be rescued after transplantation with hematopoietic progenitors obtained *in vitro* from autologous iPSC. This was achieved after correction of the human sickle hemoglobin allele by gene-specific targeting.

Fig. 5. Studies performed to validate the therapeutic potential of iPSC.

Also in mice, Xu et al. cured hemophilia by transplantation of cells that were generated from murine wild-type iPSC. These murine experiments suggest that human iPSC can be utilized for regenerative and therapeutic applications (Xu et al., 2009). Most recently, patientspecific iPSC have been established. Raya et al. reprogrammed dermal fibroblasts and/or epidermal keratinocytes of Fanconi anemia patients to generate iPSC, which were genetically corrected with lentiviral vectors encoding FANCA or FANCD2, to obtain hematopoietic progenitors of the myeloid and erythroid lineages that are phenotypically normal, that is, disease-free (Raya et al., 2009). Similar strategies were performed to correct the Hurler syndrome (Tolar et al., 2011) and for the production of macrophages from iPSCs which were resistant to HIV infection (Kambal et al., 2011).

Searching for the Key to Expand Hematopoietic Stem Cells 229

Fig. 6. Chemicals used to enhance reprogramming or to replace core reprogramming

Differentiation of iPSC into hematopoietic lineage have been achieved using a combination of specific cytokines and growth factors (Sakamoto et al., 2010) and have already demonstrated from both mouse and human iPSC (Lengerke et al., 2009; Niwa et al., 2009; Woods et al., 2011). However, the number of cells obtained in view of therapeutic use is still insufficient and small molecules that might expand the production of hematopoietic cells have yet to be found. Studies in this direction are beginning to emerge. For example, Wnt signaling, in particular WNT3a, mediates the stimulation of hemoangiogenic cell development and increase hematopoietic differentiation from ESC and iPSC (Wang and Nakayama, 2009; Wang et al., 2010). However, the conditions to generate human HSC capable of robust, long-term, multilineage engraftment from iPSC

The ex-vivo expansion of HSC represents a promising approach to obtain large enough quantities for therapeutic intervention in cell and gene therapy protocols. Derivatives of hESCs and iPS cells are also expected to be employed as *de novo* HSC source for therapeutic settings. However, as described in the previous section, practical and ethical issues must be settled before clinical practice can begin. In both cases, the chemical biology approach using small molecules as tools or drugs holds unquestionably greater promise in the outcome of

Even though a few molecules are being tested in clinical assays, the ideal soluble factor that enables to increase the number of rare HSC during the *ex vivo* growth culture without limiting their regeneration capacities has yet to be found. Most attempts have been unsuccessful because i) suitable expansion *in vitro* has been mostly correlated with loss of

**4.2.3 Obtaining HSC from induced pluripotent stem cells** 

factors.

are still hoped for.

**5. Conclusion** 

the final goal.

The enthusiasm surrounding the clinical potential of iPSC is tempered by key issues regarding their safety, efficacy, and long-term benefits. Fully realizing the biomedical potential of iPSC in a clinical setting will require addressing certain limitations inherent to the process. First, need to find alternative strategies to remove non-viral or non-integrative vectors to overcome their potential deleterious effects. Although expression of the exogenous reprogramming factors is eventually silenced during iPSC cell generation, there is a significant risk of tumorigenesis if these exogenous genes are inadvertently reactivated. Second, it will be essential to increase the number of cells with specific phenotype. Third, it will be necessary to improve the efficiency of reprogramming (0.001–3% of cells are reprogrammed) since this is a slow and inefficient process (Jaenisch and Young, 2008; Nakagawa et al., 2008; Wernig et al., 2008).

### **4.2.2 Improving reprogramming somatic cells to pluripotency**

To overcome the potential deleterious effects of viral vectors or oncogenes, and to improve the efficiency of the process toward a potential clinical application, a powerful alternative is offered by using small molecules. Several small molecules were reported to improve the reprogramming process by lowering the epigenetic barrier to initiate pluripotency (for reviews see Feng et al., 2009; Lyssiotis et al., 2011). Consistent with this notion, small molecules that affect reorganization of chromatin architecture, a rate limiting step during the reprogramming of a somatic genome, have been identified (Blelloch et al., 2006; Hochedlinger and Jaenisch, 2006; Huangfu et al., 2008a). In particular, the HDAC inhibitor VPA was shown to strongly increase reprogramming efficiency in the absence of c-Myc in both mouse and human cells and to allow 2-factor reprogramming (Oct4 and Sox2) of human fibroblasts in the absence of Klf4 and c-Myc (Huangfu et al., 2008b). Other epigenetic regulators such as BIX01294, a G9a histone methyltransferase inhibitor; BayK8644, an L-type calcium channel agonist and the two DNA methyltransferase inhibitors, AzaC and RG108 (summarized in Fig. 6), substantially increased reprogramming efficiency (Lukaszewicz et al., 2010; Shi et al., 2008).

The low efficiency of reprogramming (Hong et al., 2009) might also result from the accumulation of ROS (Parrinello et al., 2003). Consistent with this, Esteban et al. found that vitamin C strongly increases the reprogramming efficiency (Esteban et al., 2010). This is in line with the study reporting that hypoxic conditions improve the efficiency of iPSC production generated from mouse or human somatic cells (Yoshida et al., 2009). Cotreatment with VPA synergizes this effect. Other molecules including the MEK inhibitor PD0325901, the GSK3 inhibitor CHIR99021 combined with tranylcypromine, kenpaullone, SB-431542, and the TGF- signaling inhibitor called RepSox (Ichida et al., 2009) were reported to enhance reprogramming or to replace viral vectors or oncogenes (Li and Ding, 2009; Pan and Thomson, 2007).

With the continued use of high-throughput screening to identify more chemicals that could assist in reprogramming, we may be closer to the goal of using a chemical-only cocktail to reprogram somatic cells to iPSC. These pluripotency gene activators may be then used in combination with specific differentiation modulators to achieve the production of the desired cell type.

The enthusiasm surrounding the clinical potential of iPSC is tempered by key issues regarding their safety, efficacy, and long-term benefits. Fully realizing the biomedical potential of iPSC in a clinical setting will require addressing certain limitations inherent to the process. First, need to find alternative strategies to remove non-viral or non-integrative vectors to overcome their potential deleterious effects. Although expression of the exogenous reprogramming factors is eventually silenced during iPSC cell generation, there is a significant risk of tumorigenesis if these exogenous genes are inadvertently reactivated. Second, it will be essential to increase the number of cells with specific phenotype. Third, it will be necessary to improve the efficiency of reprogramming (0.001–3% of cells are reprogrammed) since this is a slow and inefficient process (Jaenisch and Young, 2008;

To overcome the potential deleterious effects of viral vectors or oncogenes, and to improve the efficiency of the process toward a potential clinical application, a powerful alternative is offered by using small molecules. Several small molecules were reported to improve the reprogramming process by lowering the epigenetic barrier to initiate pluripotency (for reviews see Feng et al., 2009; Lyssiotis et al., 2011). Consistent with this notion, small molecules that affect reorganization of chromatin architecture, a rate limiting step during the reprogramming of a somatic genome, have been identified (Blelloch et al., 2006; Hochedlinger and Jaenisch, 2006; Huangfu et al., 2008a). In particular, the HDAC inhibitor VPA was shown to strongly increase reprogramming efficiency in the absence of c-Myc in both mouse and human cells and to allow 2-factor reprogramming (Oct4 and Sox2) of human fibroblasts in the absence of Klf4 and c-Myc (Huangfu et al., 2008b). Other epigenetic regulators such as BIX01294, a G9a histone methyltransferase inhibitor; BayK8644, an L-type calcium channel agonist and the two DNA methyltransferase inhibitors, AzaC and RG108 (summarized in Fig. 6), substantially increased reprogramming efficiency (Lukaszewicz et al., 2010; Shi et al.,

The low efficiency of reprogramming (Hong et al., 2009) might also result from the accumulation of ROS (Parrinello et al., 2003). Consistent with this, Esteban et al. found that vitamin C strongly increases the reprogramming efficiency (Esteban et al., 2010). This is in line with the study reporting that hypoxic conditions improve the efficiency of iPSC production generated from mouse or human somatic cells (Yoshida et al., 2009). Cotreatment with VPA synergizes this effect. Other molecules including the MEK inhibitor PD0325901, the GSK3 inhibitor CHIR99021 combined with tranylcypromine, kenpaullone, SB-431542, and the TGF- signaling inhibitor called RepSox (Ichida et al., 2009) were reported to enhance reprogramming or to replace viral vectors or oncogenes (Li and Ding,

With the continued use of high-throughput screening to identify more chemicals that could assist in reprogramming, we may be closer to the goal of using a chemical-only cocktail to reprogram somatic cells to iPSC. These pluripotency gene activators may be then used in combination with specific differentiation modulators to achieve the production of the

Nakagawa et al., 2008; Wernig et al., 2008).

2008).

2009; Pan and Thomson, 2007).

desired cell type.

**4.2.2 Improving reprogramming somatic cells to pluripotency** 

Fig. 6. Chemicals used to enhance reprogramming or to replace core reprogramming factors.
