Induced Pluripotent Stem Cells as a Tool for Modeling Hematologic Disorders and as a Potential Source for Cell-Based Therapies

The breakthrough in human induced pluripotent stem cells (hiPSCs) has revolutionized the field of biomedical and pharmaceutical research and opened up vast opportunities for drug discovery and regenerative medicine, especially when combined with gene-editing technology. Numerous healthy and patient-derived hiPSCs for human disease modeling have been established, enabling mechanistic studies of pathogenesis, platforms for preclinical drug screening, and the development of novel therapeutic targets/approaches. Additionally, hiPSCs hold great promise for cell-based therapy, serving as an attractive cell source for generating stem/progenitor cells or functional differentiated cells for degenerative diseases, due to their unlimited proliferative capacity, pluripotency, and ethical acceptability. In this review, we provide an overview of hiPSCs and their utility in the study of hematologic disorders through hematopoietic differentiation. We highlight recent hereditary and acquired genetic hematologic disease modeling with patient-specific iPSCs, and discuss their applications as instrumental drug screening tools. The clinical applications of hiPSCs in cell-based therapy, including the next-generation cancer immunotherapy, are provided. Lastly, we discuss the current challenges that need to be addressed to fulfill the validity of hiPSC-based disease modeling and future perspectives of hiPSCs in the field of hematology.


Introduction
Over a decade ago, human cellular reprogramming from adult somatic cells to pluripotent state, so-called human induced pluripotent stem cells (hiPSCs) was successfully discovered [1,2]. Due to their unique properties of unlimited self-renewal and the ability to differentiate into all cell types of three primordial germ layers, termed 'pluripotency', hiPSCs offer numerous opportunities for a wide range of research and clinical applications. In particular, the groundbreaking discovery of hiPSCs has revolutionized the field of biomedical research and the study of human physiology and diseases, which ideally require that the studied cells and tissues are obtained directly from patients. In some cases, primary human cells or tissues are not easily accessible, e.g., those from the brain and heart; hence, research with certain human materials is limited. One of the merits of using hiPSCs is the potential to overcome this limited donor accessibility that, together with differentiation techniques, can establish an appropriate model system for developmental and disease studies at various pathological events from an early stage of disease progression, otherwise unavailable in primary cells. Importantly, hiPSCs that are derived from patients other studies, additional genes or enhancers, either small molecules or chemicals, were applied to increase reprogramming efficiency. For example, undifferentiated embryonic cell transcription factor 1 (UTF1), T-box3 transcription factor TBX3, siRNA or shRNA against p53, histone deacetylase inhibitor valproic acid (VPA), and vitamin C have been reported to enhance hiPSC generation [15][16][17][18][19][20][21]. On the other hand, several studies have attempted to simplify the reprogramming process by minimizing reprogramming factors, e.g., from OSKM to only OCT4 and SOX2 [20] or OCT4 alone [22], albeit with lower efficiency.
The efficiency of hiPSC generation is influenced greatly by the delivery method of reprogramming factors. Generally, the reprogramming efficiency of human somatic cells to hiPSCs even with viral-based method is limited, i.e., up to~2% [23]-only a small number of transduced cells can complete molecular events and become fully reprogrammed iPSCs [24]. Concerning their safety profile, delivery systems are generally classified into two main types based on their effects on genomic integration of transgenes. Integration systems, such as retroviral and lentiviral-based vectors, were used in early studies to establish mouse and human iPSCs from various cell types [1,2,6,25]. These viral inductions provide stable transgene expression and more efficient iPSC derivation, when compared to plasmid transfection and, as predicted, non-integration systems [24]. Notably, the lentiviral transduction has advantages over the retroviral transduction since it can infect both dividing and nondividing cells, giving higher reprogramming efficiency. However, the risks of insertional mutagenesis and transgene reactivation remain their drawbacks. Shortly thereafter, several integration-free systems were developed to facilitate future clinical applications. These integration-free methods can be subdivided into: (i) viral-based vectors, e.g., adenovirus and Sendai virus (SeV); (ii) DNA-based vectors, e.g., episomal plasmids (Epi) and minicircle DNA; (iii) self-excision vectors, e.g., piggyBac transposon; (iv) modified mRNA; and (v) recombinant proteins. A previous study performed systemic evaluation on the most commonly used integration-free methods: SeV, Epi, and mRNA, and reported different efficacies and advantages among them. SeV reprogramming offers relatively high efficiency (0.01-1%) and is a reliable method [26]; it requires a smaller workload and was validated in multiple cell types, including fibroblasts and blood cells, e.g., peripheral blood mononuclear cells (PBMCs), CD34 + cells, and T cells for both researchand clinical-grade iPSC generation [27][28][29][30][31][32]. For Epi reprogramming, it is also highly reliable and feasible to generate clinical hiPSCs under current good manufacturing practice (cGMP), making it suitable for transition to clinical use [26,32]. mRNA reprogramming offers the highest efficiency (up to 4%) and complete absence of integration [33,34]; however, this technique is laborious mainly due to the requirement of multiple transfections. It is worth noting that the cell source used for mRNA reprogramming was mainly fibroblasts; hence, the information in blood cells is very limited [26,35]. Additionally, numerous reports have described the pivotal role of specific microRNAs (miRNAs) in reprogramming. Several groups of miRNAs have been identified as being uniquely expressed in ESCs when compared to differentiated cells. Among them, the miR-302/367 cluster has gained much attention as its expression was highly expressed in the early embryo and then rapidly decreased upon differentiation. miR-302/367 was shown to greatly improve the somatic cell reprogramming mediated by OSKM and OSK [36,37]. Anokye-Danso and colleagues reported that miR-302/367 was able to reprogram human fibroblasts even without other exogenous transcription factors [38]. However, there is still an unmet demand for optimal protocols for mRNA or miRNA reprogramming in blood cells. Although hiPSCs can be theoretically derived from any specific cell types, the starting cells with reported limited reprogramming efficiency include terminally differentiated blood cells, i.e., T cells, B cells, macrophages, and granulocytes [39].
The reprogramming process also involves culture conditions. Initially, a feeder system, such as inactivated mouse embryonic fibroblasts (MEFs), was used upon reprogramming to generate and maintain hESCs, and was later applied to hiPSCs. Subsequently, a feeder-free condition was established to make hiPSCs more applicable for clinical use. There have been different studies on the combinations of extracellular matrices and culture media to create a xeno-free condition for reliable propagation of hiPSCs, for example, the culture of hiPSCs on a vitronectin-or recombinant laminin-511 E8 fragment (rLN511E8)-coated surface instead of Matrigel coating in defined xeno-free media, e.g., E8 or mTeSR1, to create a xeno-free condition for clinical compliance [40][41][42]. It should be noted that optimization of these culture conditions is required for the cell type-to-cell-type basis.
In an hiPSC-based disease model, particularly for genetic disorders, the type of starting cells is a major consideration. In hereditary genetic diseases where all cells carry the disease-causing mutations, the choice of cell types is more diverse compared with acquired genetic disorders. The cell type of the latter is restricted because acquired somatic mutations only affect certain non-germline cells, for example, only hematopoietic cells are affected in acquired hematologic diseases. Thus, the choice of starting cells is greatly important, as it represents genetic information or disease-associated mutations of the donor cells. Furthermore, identifying the disease phenotypes is also the key advantage of iPSC-based disease modeling. This strategy requires an appropriate cell differentiation protocol to produce disease-relevant cells and the development of well-control cell line used to compare the molecular and cellular phenotypes. As mentioned above, each hiPSC line may retain epigenetic memory from incomplete reprogramming. In this case, it causes line-to-line variation that may impact the data interpretation and differentiation process. However, this hurdle can be overcome by using genome editing technology to create isogenic pairs, by either correction of the disease-associated mutations or introduction of mutation in normal control iPSCs (see the following sections).

Hematopoiesis and Hematopoietic Differentiation from hiPSCs
The differentiation potential of hiPSCs into any clinically relevant cell types offers great promise and opportunities for cell-based therapy and regenerative medicine, and also offers a unique chance to model human disease pathology and drug screening. Several attempts have been made to develop blood cells from hiPSCs based on the findings on embryonic hematopoiesis studied in ESCs, which were established two decades ago, or animal models. Generally, developmental hematopoiesis occurs as spatiotemporally overlapping waves and can be categorized into three main waves based on the characteristics: (i) primitive hematopoiesis, (ii) pro-definitive hematopoiesis, and (iii) definitive hematopoiesis. The first wave appears in the yolk sac and gives rise to primitive blood cells, including nucleated erythrocytes, megakaryocytes, and macrophages, but lacks lymphoid potential [43][44][45]-it provides the growth needs and innate defense mechanisms for the embryo (Figure 1).
The second wave is where adult-like blood cells emerge. In this wave, the erythromyeloid progenitors (EMPs) originate from the hemogenic endothelium (HE) in the yolk sac and develop into erythrocytes, megakaryocytes, and myeloid lineages and subsequently migrate to the fetal liver. The second wave also produces lympho-myeloid progenitors (LMPs), which are undetectable in the primitive wave [46][47][48][49]. However, the cells in this transient wave have no long-term repopulating potential and cannot reconstitute hematopoiesis. Following this program, the first definitive hematopoietic stem cells (HSCs) emerge in the aorta-gonad-mesonephros (AGM) region, a potent hematopoietic site within the mammalian embryo body, through the endothelial-to-hematopoietic transition (EHT) process and localize in the fetal liver, where they propagate before homing to the bone marrow [50][51][52]. These definitive HSCs have multilineage differentiation and long-term engraftment potential as well as the capability of life-long hematopoiesis in adult life [52,53]. Hence, the establishment of definitive HSCs in vitro has gained much attention in the field of hematology and is a promising approach for future regenerative medicine. The process of hematopoiesis is associated with a dynamic regulation of various factors such as transcription factors, signaling pathways, microenvironment called the niche, and interaction between the cells. Understanding these factors enables the efficient production of blood lineages and can be, ultimately, translated to clinical used. It has been shown that the production of blood lineage from hPSCs can recapitulate early blood development during ontogeny. Several studies have demonstrated different strategies for in vitro hematopoietic differentiation from hiPSCs. The common process of these procedures begins with the differentiation of hiPSCs to exit the pluripotent state and move toward intermediate cell states such as the mesoderm and hemogenic endothelium, followed by a terminal differentiation into specialized blood cells of interest ( Figure 2  Previous studies have demonstrated that bone morphogenetic protein 4 (BMP4), fibroblast growth factor 2 (FGF2), WNT/β-catenin, and Activin/Nodal play critical roles for mesoderm formation [54][55][56][57]. Thus, specific hematopoietic growth factors and cytokines, i.e., BMP4, Activin A, CHIR99021 (GSK-3 inhibitor), SB-431542 (TGFβ inhibitor), vascular endothelial growth factor (VEGF), FGF2, stem cell factor (SCF), thrombopoietin (TPO), interleukin (IL)-3, and IL-6, were added throughout the differentiation process to induce HSC specification. In addition, lineage specification markers have been described to specify different cell populations occurring upon hematopoietic differentiation. For example, KDR, apelin receptor (APLNR), and platelet-derived growth factor receptor alpha (PDGFRα) have been used for the emergence detection of the mesoderm [58]. Some studies have used the marker KDR + CD235a − to identify mesodermal cells that give rise to a definitive hematopoietic program [59]. The hematopoietic induction system could be classified into three approaches: (i) the formation of embryoid bodies (EBs), of which the cells form a 3D structure that allows cell-cell interaction, mimicking the spatial organization of the embryo; (ii) co-culture with stromal feeder cells such as OP9 cells, which provides an intimate cell contact with secreting factors that promote cell proliferation and differentiation; and (iii) a 2D monolayer culture on an extracellular matrix (ECM)-coated dish that together with hematopoietic cytokine cocktails could generate a xeno-free culture [58,60,61]. Although early studies demonstrated mostly primitive HSC production from hiPSCs, recent studies have revealed improved methods by which to develop hiPSC-derived definitive HSCs, which are characterized by the presence of adult globin expression in erythrocytes, multilineage differentiation including T lymphocytes, and long-term engraftment ability in immunocompromised mice models [60,62]. Transcription factors, e.g., HOXA9, RUNX1a, and MLL-AF4, were shown to enhance hematopoietic commitment, multilineage differentiation, and engraftable HSPCs [63][64][65][66]. Although MLL-AF4 was sufficient to re-specify iPSCs into long-term engrafting CD34 + CD38 − CD90 + CD49f + HSPCs in immunodeficient mice, reconstituting long-term B, T, erythroid, and myeloid cells, MLL-AF4-induced HSPCs were prone to leukemic transformation during the long-term posttransplant period [66]. The efficient production of definitive HSCs would offer more clinical potential for transplantation and the production of functional bloods for cell replacement therapy and immunotherapy, yet optimal differentiation protocols remain to be elucidated. Notably, in vivo HSC differentiation from hiPSCs via teratoma formation has also been proposed to generate functional HSCs and long-term reconstitution, as this system was suggested to provide the microenvironment that resembled HSC niche [67][68][69]. However, it should be noted that specific cell types obtained from teratomas require thorough isolation prior to use, to avoid the contamination of other cells, and that this system was performed in animals; hence, this approach might be impractical for clinical use.
hiPSC-derived HSCs/HSPCs have the potential to develop further into various blood cell types: red blood cells (RBCs), megakaryocytes (MKs), myeloid cells, and lymphoid cells via terminal differentiation. For erythropoiesis, studies of hiPSC-derived RBCs have been ongoing for many years. It has been noted that RBCs differentiated from iPSCs are likely to resemble yolk sac primitive/definitive and fetal liver EMP-erythropoiesis because they have a low enucleation potential and embryonic/fetal globin expression. Several attempts have been made to overcome these hurdles by adding small molecules, e.g., inhibitor VIII (specific GSK-3β inhibitor) and 3-isobutyl-1-methylxanthine (IBMX) (cAMP inhibitor) [70], in the culture system, pre-selection definitive erythroid population [71], and manipulation of specific genes involved in erythropoiesis, e.g., c-MYB, BCL11A, KLF1, and SH2B3 [72][73][74][75]. However, the enucleation rate remains slightly improved and the hemoglobin content in hiPSC-derived RBCs is partially expressed adult globin [76][77][78]. In terms of cell expansion, hiPSC-derived RBCs showed a lower expansion rate in comparison to RBCs derived from cord blood/mobilized peripheral blood [79]. Therefore, the development of RBCs from hiPSCs requires more extensive studies to improve the quality and quantity of produced RBCs. For megakaryopoiesis, hiPSCs have been differentiated into megakaryocyte lineages with various improved methods, i.e., supplementation of platelet stimulation by TPO or agonist to c-MPL (TPO receptor) in the culture, downregulation of c-MYC by specific compound iBET151, and exogenous expression of GATA1, FLI1, and TAL1 in hiPSCs, to increase MK expansion and maturation [80][81][82]. The major challenge of hiPSC-derived MKs is the generation of sufficient functional pro-platelets that might be used in clinical applications. It has been shown that culture conditions with shear stress enhanced platelet production from hiPSC-derived MKs and could be further scaled up in a bioreactor [83,84]. This progress paves the way to produce high-yield and functional platelets in vitro, and to utilize hiPSC-derived platelets in blood transfusion. The first-in-human clinical trial for hiPSC-derived platelets was demonstrated in an aplastic anemia patient who suffered from platelet transfusion refractoriness. The study used autologous iPSC-derived platelets to treat the patient and there was no evidence of an alloimmune response [85]. For the myeloid lineage, hiPSC-derived macrophages or monocytes have been more widely studied in this field due to their similarity to embryonic development that produces tissue-resident macrophages [86]. Furthermore, hiPSC-derived macrophages have been described as having similar transcriptomic profile to tissue-resident macrophages, which are normally difficult to access [87]. Previous studies demonstrated the capability of the differentiation of hiPSCs toward monocytes and functional macrophages, and this could extend to a large-scale production [88]. These systems can be applied for disease modeling, cellular therapy, and drug screening. For lymphoid cells, scholars have reported the generation of hiPSC-derived T cells through co-culturing on delta-like (DL)-1 or DL-4 expressing OP9 cells and in vivo teratoma formation [89][90][91]. A recent study established T cells derived from T cell receptor (TCR)-engineered iPSCs using a feeder-free protocol [92], with the aim of their utilization in alternative cell-based immunotherapy. Collectively, the generation of HSCs and lineage-specific blood cells from hiPSCs is a complex process of developmental hematopoiesis. It relies on the coordination of multiple factors to achieve efficient production. More extensive studies may fulfill the knowledge in this field and make them more applicable in a clinical setting.

Genetic Disease Modeling with Patient-Specific iPSCs
The emergence of hiPSC technology has contributed to rapid progress in the study of various human diseases, particularly those that have no previous physiologically relevant experimental models. In the past, animal models, frequently mouse models, human primary cells, and immortalized cell lines represented standard experimental tools in basic research to delineate disease pathophysiology, as well as in pharmaceutical research and the development of new drugs. However, mouse models do not always fully recapitulate human phenotypes and biological responses due to species differences, and cell culture models are limited to certain diseases, mainly due to insufficient expandable and/or difficult-to-access cell sources, e.g., cardiomyocytes, neuronal cells, pancreatic cells, and hepatocytes. The use of hiPSCs has solved these hurdles, and due to their property of being able to be maintained without limit in cultures, and their ability to differentiate into virtually any cell type, they have gained much attention in the modeling of diseases with complex genetic defects. Furthermore, hiPSCs can be directly generated from the reprogramming of patients' somatic cells, allowing a patient's genetic identity to be preserved in hiPSCs. To date, various cell types differentiated from the patient-specific iPSCs have been shown to recapitulate the disease phenotypes under appropriate conditions, reflecting the correlation of genotypes to cellular phenotypes. In the case of acquired genetic disorders, where disease-causing mutations are only presented in specific tissues, hiPSCs can be established from those normal somatic cells in parallel, thereby providing an opportunity to create normal hiPSCs from the same patient to be used as an isogeneic control. The combination with recent gene-editing techniques, such as RNA-guided clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9), further increases the value of hiPSCs in the modeling of genetic diseases, as well as in clinical cell-based therapy ( Figure 3). In this section, we summarize the example studies using hiPSCs for genetic disease modeling for various hereditary and acquired hematologic disorders (Table 1), discuss the major concepts, and demonstrate their potential applications in drug discovery.  Patient-specific iPSCs are a powerful tool for elucidating disease mechanisms and drug screening. Likewise, gene-editing technology can be employed to correct mutations in patient-specific iPSCs, making them a promising cell source for large-scale cell production for cell-based therapy. iPSC, induced pluripotent stem cell; HSC, hematopoietic stem cell.

Hereditary Hematologic Disorders
Significantly, hiPSCs have been widely used in the modeling of hereditary hematologic disorders in which disease-associated mutations can be passed from the affected individual to the next generation, also known as inherited disorders. The most common and well-known hereditary hematologic disorders are hemoglobinopathies that are caused by mutations in the globin genes. For example, β-thalassemia is characterized by the mutations in β-globin gene (HBB), resulting in a reduction in or an absence of β-globin expression. It has been shown that the hiPSC-derived erythroid cells from patients with various forms of β-thalassemia displayed low levels of the HBB gene and hemoglobin protein expressions. After genetic modifications, i.e., by homologous recombination or an engineered nucleases-based system, such as zinc-finger nucleases (ZFNs), transcription activator-like effector nucleases (TALENs), and CRISPR/Cas9, corrected iPSCs could be differentiated into erythroid cells with improved β-globin production [93][94][95][96][97][98][99]102,143]. Hemoglobin E (HbE)/β-thalassemia represents the major genotype of those patients affected with severe β-thalassemia worldwide and is the most common form of β-thalassemia found in adults in Southeast Asia. In Thailand alone, at least 3250 new cases of HbE/β-thalassemia are expected annually (1 in 180 births). In HbE/β-thalassemia, one allele (β 0 ) produces no β-globin chain, and the other allele (β E ) produces an HbE globin chain resulting from nucleotide substitution at codon 26 (GAG → AAG, glutamic acid to lysine). Our group used the CRISPR/Cas9 system and a single-stranded DNA oligonucleotide (ssODN) donor template to efficiently correct the HbE mutation in HbE/β-thalassemia patient-derived hiPSCs in one step (2.9% homology-directed repair (HDR)). The corrected iPSCs, which are β-thalassemia heterozygotes, are capable of differentiation into CD34 + CD43 + HSPCs and subsequent erythroid cells that express normal HBB and β-globin [93]. The feasibility of genetic correction makes these hiPSCs more attractive for cell therapy by using corrected hiPSC-derived HSPCs for autologous transplantation in the future. Notably, the state of iPSCs has been demonstrated to affect gene correction efficiency-the naïve iPSC reprogrammed from patient with β-thalassemia (41/42, -TCTT deletion) appeared to have higher gene-editing efficiency than those conventional primed iPSCs [144].
Another type of hereditary disorder is hemolytic anemia. In this case, patients suffer from RBC destruction before the end of the cells' normal life span of about 120 days, and hence the patients require blood transfusion or other supportive treatments. A previous study showed that erythroid cells derived from patients' iPSCs with pyruvate kinase deficiency (PKD), a rare metabolic blood disease caused by the mutation in PKLR gene, exhibited energetic imbalances such as decrease in adenosine triphosphate (ATP) levels and metabolites in the glycolysis pathway; the study also showed that the disease phenotypes were restored after knocking in the PKLR gene via TALEN-mediated homologous recombination [103]. Hereditary hemolytic anemia could also be caused by mutations in Krüppel-like factor 1 (KLF1), a transcription factor controlling almost all aspects of erythropoiesis, although mutations in KLF1 can vary and lead to different phenotypes ranging from mild to severe phenotypes. A recent study demonstrated disease modeling using hiP-SCs derived from a patient with type IV congenital dyserythropoietic anemia (CDA), a rare and severe disease caused by a heterozygous mutation on the second zinc finger of KLF1 (c.973G > A, p.E325K). The CDA-iPSCs-derived erythroid cells displayed multinucleated morphology, a lack of CD44 expression, and KLF1 target gene dysregulation. Moreover, this study suggested the underlying mechanisms of the p.E325K mutation, which involve the disruption of cell cycle regulators, the cell membrane, and enzyme abnormalities [104]. Recently, our group has generated an hiPSC line from a pediatric patient with severe hemolytic anemia carrying compound heterozygote mutations in KLF1 (G176RfsX179 and A298P) [145]. Subsequently, the hiPSCs were differentiated toward the erythroid cells. The KLF1-iPSC-derived erythroid cells exhibited relatively low proliferation, poor differentiation, decreased erythroid related-gene expression, and cell-cycle dysregulation when compared to normal hiPSCs, the phenotypes of which could be partially restored after CRISPR/Cas9 gene editing of the G176RfsX179 mutation.
In addition, hiPSCs have been utilized to model immune system disorders, including severe combined immunodeficiency (SCID) and chronic granulomatous disease (CGD). SCID is a rare disease classified as primary immunodeficiency caused by mutations in multiple genes, e.g., IL2RG, JAK3, ADA, and RAG1/RAG2, which are involved in the development of functional T and B cells in the immune system. Several hiPSC modelings of SCID from different mutations have been generated. Previous studies showed that T cells derived from SCID patients with JAK3 and IL2RG mutations exhibited blockage in early T cell development or were unable to produce functional lymphocytes. These defective phenotypes were rescued after gene correction [118,119]. Moreover, recent studies used an SCID-iPSC model to investigate the stage of T cell development and T cell receptor (TCR) rearrangement [89,120]. CGD, which is caused by mutations in the CYBB gene encoding NADPH oxidase that is associated with reactive oxygen species (ROS) production in phagocytes, resulted in recurrent infections. The hiPSCs derived from patients with CGD demonstrated the disease-relevant phenotypes resembling those in the patients-that is, the lack of ROS production in CGD-iPSC-derived neutrophils or macrophages [110,111]. Importantly, the defective phenotypes were restored after correction of disease mutations in CGD-iPSCs using different approaches [112][113][114][115][116]. A recent study employed the CRISPR/Cas9 system to first generate patient-like hiPSCs by introducing the p47-∆GT mutation into normal hiPSCs and subsequently mediating gene editing of the introduced mutation. Granulocytes differentiated from such corrected hiPSCs could restore gene expression and express functional NADPH oxidase activity with bacteria-killing capacity [117].

Acquired Hematologic Disorders
Acquired hematologic disorders are not inherited, and disease-causing mutations are acquired and restricted only in certain hematopoietic cells, resulting in cell heterogeneity in patients. Disease-specific hiPSCs can be then generated from the abnormal clones, while normal hiPSCs that do not carry genetic mutations can be generated from unaffected tissues to serve as germ-line controls in experiments or can be used as autologous cells for therapeutic purposes. Hematologic malignancies are a good example of acquired hematologic disorders, which contain very heterogeneous subpopulations. Samples collected from the patients' bone marrow or PBMCs could be a mixture of normal cells and premalignant and malignant clones, depending on the disease progression, remission, and administered therapies. It is believed that generated hiPSCs from different clones can be used to study clonal evolution that is a critical event in cancer development. However, efforts to develop hiPSC models of hematologic malignancies are thus far limited to myeloid malignancies, namely, myeloproliferative neoplasms (MPN), myelodysplastic syndromes (MDS), MDS/MPN overlap syndromes, and acute myeloid leukemia (AML), mainly due to the difficulty of generating lymphoid lineages from hiPSCs.
MDS, earlier described as preleukemia or smoldering leukemia, is a bone marrow failure syndrome characterized by impaired hematopoiesis. Previously, hiPSCs were derived from the hematopoietic cells of MDS patients with chromosomal 7q deletions (del7q) in parallel with isogenic normal hiPSCs from hematopoietic cells. del7q-iPSCs displayed hematopoietic defect phenotypes after differentiation, which could be rescued after chr7q dosage correction. This system was extended to identify the candidate diseasespecific haplo-insufficient genes, such as EZH2 and LUC7L2, which might mediate the hematopoietic defects in MDS via phenotype-rescue screening [134]. To explore the disease stage transition, a panel of hiPSCs capturing distinct disease stages from preleukemia, low-risk MDS, high-risk MDS, and secondary AML, was created. Upon differentiation, the disease-stage-specific iPSCs could recapitulate the graded severity of cellular phenotypes and/or stage specificity observed in the patients. This study also demonstrated the use of a CRISPR/Cas9 system to induce both disease progression and reversal via genetic correction or introduction of mutations, which may aid in the mechanistic understanding of myeloid transformation and drug testing for stage-specific therapeutic interventions [135].
Another study also used an hiPSC model of MDS derived from single premalignant cells with a partial complement of mutations in an investigation of acquired genetic changes and their functional consequences on disease progression from MDS-iPSC subclones. Using this approach, the study was able to identify the events that occurred during leukemic progression, for example, SF3B1 mutations, in concert with epigenetic mutations that could disturb mitochondrial function, leading to damaged mitochondria and resulting in apoptosis and impaired erythropoiesis [146]. A recent study developed a model of clonal evolution of AML by sequentially introducing the associated mutations, i.e., ASXL1, SRSF2, and NRAS, using CRISPR/Cas9 technology in hiPSCs, and characterized the transcriptional program and signaling pathways upon disease development using transcriptomic and chromatin analyses. This study also suggested a novel therapeutic approach that targets early AML via an inhibition of inflammatory signaling, which is an early and persistent event in leukemogenesis [147].
Additionally, hiPSCs from AML patients with MLL rearrangements have been successfully established, in which the mutations were preserved in the derived AML-iPSCs, while leukemic DNA methylation and the gene-expression profile were reset after reprogramming. Upon hematopoietic differentiation, the leukemic properties were reactivated as indicated by an aberrant myeloid-restricted phenotype and aggressive myeloid leukemia after transplantation into immunodeficient mice. In addition, distinct AML-iPSC subclones were used to identify target therapy at a specific stage of AML [136]. hiPSC models of chronic myeloid leukemia (CML) have been developed displaying disease phenotypes such as impaired hematopoietic differentiation, resistance to tyrosine kinase inhibitors (TKIs), and heterogeneity of the clones [137]. These models were used to gain more extensive insight into the mechanisms of TKI resistance in CML stem cells and were evaluated as therapeutic candidates that could be used to prevent disease recurrence [138][139][140][141].
Polycythemia vera (PV) and essential thrombocytothemia (ET) are BCR-ABL1 negative MPNs, involving an increased red cell mass and clonal platelet overproduction, respectively. JAK2 is the most common target of driver mutations with frequencies of approximately 98% in PV and 50-60% in ET. Previous studies generated iPSCs from cells of MPN patients carrying the JAK2V167F mutation, usually detectable in both PV and ET, and JAK2 exon12 mutations found in PV, and subsequently differentiated them toward a hematopoietic lineage. JAK2exon12-iPSCs, but not JAK2V617F-iPSCs, exhibited an increased erythropoiesis resembling the pathophysiology in PV patients. JAK2V617F-iPSCs did not exhibit a significant increase in erythroid cell proliferation or differentiation, most likely because the studied JAK2V617F expression was more consistent with the heterozygous JAK2V617F mutation in ET patients, where the endogenous JAK2 gene was still present [148,149]. The utilization of different types of JAK2 mutations in patient-derived iPSCs or generated mutant iPSCs for the discovery of pathogenesis and candidate therapy for MPNs has been demonstrated [149,150]. Additionally, hiPSCs derived from patients with the calreticulin (CALR) gene mutation, which is usually found in ET and primary myelofibrosis patients, could reflect disease phenotypes by representing megakaryopoiesis and prominent colony-forming unit megakaryocytes (CFU-MK) [151,152].

Patient-Specific iPSCs for Drug Screening
Besides applications in studying disease mechanisms, patient-specific iPSCs have contributed greatly to therapeutic development due to their limitless availability and capacity to differentiate into relevant cell types while reflecting human disease phenotypes. Compared to conventional animal-based testing and cell-line-based compound screening, the hiPSC-based platform is a more comprehensive, powerful tool for predicting the safety and efficacy of drug responses in the human setting, thus saving time and costs in the development process. Thus far, various patient-specific iPSC models for drug screening for patients with hematologic disorders have been demonstrated.
An early example platform for drug screening for hereditary blood diseases is the use of hiPSCs derived from X-linked CGD patients as a robust model for testing enzyme re-placement therapy. The researchers utilized recombinant NOX2/p22phox proteoliposomes containing an active cytochrome b558, the membrane component of NADPH oxidase complex, and delivered it to CGD-iPSC-derived macrophages. After treatment with specific liposomes, NADPH oxidase activity was restored without toxicity, making it a potential approach for future treatment of pulmonary infection in CGD patients [153]. For Fanconi anemia (FA), a congenital disorder characterized by bone marrow failure, FA-iPSCs were used for drug screening of several compounds known to improve FA phenotypes, such as resveratrol (Sirt1 activator), danazol (synthetic androgen), and doramapimod (p38 MAPK inhibitor), based on the evaluation of the effects on hematopoietic differentiation. Of those compounds, doramapimod and tremulacin (anti-inflammatory) significantly improved the production of CD34 + CD43 + HSPCs upon FA-HSC differentiation through a mechanism that most likely involved the suppression of proinflammatory cytokines, including IFNγ, TNF, and IL-6, at the transcriptional level [154]. Alternatively, for Schwachman-Diamond syndrome (SDS), a bone marrow failure syndrome with propensity to develop MDS and AML, researchers first used the iPSC model to identify the underlying mechanisms and candidate therapeutic target before proposing potential therapeutics. Both iPSC-derived from SDS patients and engineered del(7q) iPSCs suggested the TGF-β pathway as the candidate target. After adding small molecule SD208, an inhibitor of TGF-β receptor I kinase, into the culture, the hematopoiesis of SDS-iPSCs was rescued, as shown by an increase in both the number and size of erythroid and myeloid colonies. However, similar results could not be observed for the engineered del(7q) iPSCs, thereby suggesting that hiPSCs can be used to study the distinct contributions of somatic alterations [155].
A recent study performed chemical screening on iPSCs derived from a Diamond-Blackfan anemia (DBA) patient and identified that SMER28, a small molecule modulator of autophagy, has the potential to cure DBA. SMER28 was shown to stimulate erythropoiesis in DBA-iPSCs by promoting autophagy in erythroid progenitors through autophagy factor ATG5 and also upregulated the expression of globin genes [132]. Another group also used a DBA-iPSC model and identified a new therapy, namely, eltrombopag (EPAG), a Food and Drug Administration (FDA)-approved synthetic small-molecule mimetic of TPO. EPAG partially rescued erythropoiesis by mediating intracellular iron restriction in DBA-iPSC-derived erythroid cells, which was in common with the effects of deferasirox (DFX), a clinically licensed iron chelator, on DBA-iPSCs [156].
The discovery of therapeutic drugs has also been successful in hiPSCs derived from patients with acquired blood disorders. For example, a drug-testing system was established from hiPSCs derived from a patient with chronic myelomonocytic leukemia (CMML), a clonal HSC disorder with overlapping features of MDS and MPN. In this study, the researchers used MEK inhibitor PD0325901 and Ras inhibitor salirasib to treat CMML-iPSCderived CD34 + CD43 + HPCs and observed its capacity to suppress colony formation [157]. Moreover, phenotypic screening using liposomal clodronate, which has been previously shown to deplete monocytes in a monocyte/macrophage system, was tested under this setting. The results showed that liposomal clodronate decreased the number of colonies and reduced serial replating capacity in CMML-iPSC-derived HPCs, indicating a potential drug reposition of this clinically used osteoporosis drug. The findings also recommend hiPSCs as a rational model for the identification of drug candidates for repositioning.
Drug testing for disease-stage-specific therapeutic interventions could be performed using hiPSC models. In a myeloid malignancy study, a panel of disease-stage-specific iPSCs for modeling MDS and secondary AML was established, as mentioned above. The researchers applied 5-Azacytidine (5-AzaC), a hypomethylating agent used as first-line therapy of MDS, to different MDS-iPSCs-derived HPCs and observed its selective effect on stage-specific iPSC lines, i.e., high-risk MDS was the most responsive. Moreover, they also reported that HPCs derived from a less-and a more-advanced disease stage of MDS/AML-iPSCs displayed different responses to rigosertib, a small molecule inhibitor of RAS signaling currently undergoing clinical trials for high-risk MDS, with the KRASmutated line being more responsive to the treatment [134]. Another study identified drug sensitivity on a CRISPR/Cas9-mediated MDS-iPSC model. Treatment of engineered SRSF2mutant iPSC-derived HPCs with splicing inhibitor E7107 and splicing-modulating drugs Cpd-1, Cpd-2, and Cpd-3, which are small molecule inhibitors of CDC-like kinases and serine-arginine protein kinases, exhibited a selective growth-inhibitory effect. Further screening from a library of 2000 compounds containing FDA-approved drugs, natural products, and other bioactive compounds found that niflumic acid was identified to have selective inhibition on the growth of del(7q) iPSC-derived HPCs [158].

iPSCs and Therapeutic Applications
The promise of hiPSCs lies in their far-reaching potential to serve as a starting material by which to develop a range of therapeutic cells from their capability to multiply indefinitely and the ability to generate virtually any cell types upon differentiation in vitro. The feasibility of performing genetic manipulation further advances the therapeutic applications of hiPSCs for genetic disorders. One of the major concerns for clinical-grade iPSCs is the comparability of lines derived from different individuals and in different facilities. Previously, guidelines on critical quality attributes (CQAs) and minimum testing requirements for clinical-grade iPSC lines have been outlined: (i) identity by single tandem repeat (STR) genotyping; (ii) microbiological sterility; (iii) genetic fidelity and stability by testing for residual reprogramming vectors and karyotyping; (iv) viability; (v) characterization of markers from standard hPSC panel; and (vi) potency, which is the qualitative measure of the biological activity of the cells [159]. In this section, we discuss the uses of hiPSCs as an alternative cell source for cell-based regenerative therapy for hematologic disorders and for cell-based cancer immunotherapy in both solid tumors and hematologic malignancies.

Cell-Based Therapy
In the past, treatment options for inherited or acquired hematologic disorders mostly relied on supportive treatment, including blood transfusion, medication, surgery, and chemotherapy, aiming to relieve signs and symptoms associated with the disorder. However, these approaches are unable to completely eradicate the particular genetic causes in the patients. Hematopoietic stem cell transplantation (HSCT) has become the standard of care for many patients with malignant and non-malignant hematologic disorders, as it can reestablish lifelong blood cell production, replacing affected blood cells and allowing for a definitive cure for the disorders. Still, the major challenge of HSCT is the availability of appropriate and sufficient donor cells for robust hematopoietic repopulation. HSCT is classified into allogeneic and autologous HSCT, according to whether the cell source is from healthy donors or the patients' own stem cells, respectively. The advantages of autologous over allogeneic HSCT are the lower risk of life-threatening complications, such as graft-versus-host disease (GvHD), graft failure and infections, and the relatively good tolerance in elder patients. However, autologous HSCT may require gene modification prior to transplantation. On the other hand, allogeneic HSCT may be limited by the availability of compatible donors and in the case of hematologic malignancies, autografts may be contaminated with tumor cells and patients will not benefit from the graft-versus-tumor effect observed after allogeneic HSCT. HSCs can be harvested from bone marrow aspiration, apheresis of stimulating factor-mobilized peripheral blood (the most common source), and umbilical cord blood. Although umbilical cord blood is rich in HSCs and enables greater tolerance to human leukocyte antigen (HLA) mismatching than other sources, cell numbers remain insufficient in large adult hosts; hence, it requires ex vivo expansion [160]. Considering the feasibility for gene modification and generation of HSCs, hiPSCs are a good candidate cell source for autologous HSCT. As mentioned earlier, disease-specific iPSCs can be derived from patient's somatic cells, and the disease-causing mutation can be corrected prior to HSC differentiation. In the case of acquired hematologic disease, healthy clones of hiPSCs derived from the patient can be directly used to produce autologous HSCs, which are of less concern for immune rejection. Our group generated disease-free iPSCs from the fibroblasts of a patient with paroxysmal nocturnal hemoglobinuria (PNH), whose granulocytes and red blood cells consist of a minor normal population and major population with decreased surface CD55 and CD59 [161]. After hematopoietic induction, the differentiated cells from PNH fibroblast-iPSCs expressed early hematopoietic markers (CD34 and CD43) with normal CD55 and CD59 expression similar to those of their parental cells, suggesting that PNH fibroblast-iPSCs can be a potential source of HSCs for autologous transplantation to cure PNH patients. Notably, the first clinical trial of hiPSC-based cell therapy was in patients with neovascular age-related macular degeneration (AMD) [162], where retinal pigment epithelial (RPE) cells were derived from patients' iPSCs and transplanted back as an RPE cell sheet under the retina (#UMIN000011929). After over 2 years of transplantation, the safety and efficacy were confirmed with no sign of immune rejection. Recently, another clinical trial was undertaken using an hiPSC-immortalized megakaryocyte cell line (iMKCL) mediated with c-MYC, BMI1, and BCL-XL as a source for platelet production for a patient with thrombocytopenia from aplastic anemia [163]. Collectively, hiPSCs hold great promise for both autologous HSC and iPSC-derived blood cells for cell replacement therapy.
Notably, hiPSCs are being preferentially considered for allogeneic transplantation due to the fact that personalized hiPSC-based cell products require high cost, a complex cell facility, and are time consuming, making them impractical for acute conditions. One of the current strategies to circumvent immune rejection in allogeneic hiPSC applications is by generating hiPSCs from various HLA homozygous donors, which involves matching of the worldwide population at the major loci and stocking these cells in biobanks. However, this approach is very challenging for donor eligibility as it requires a large number of HLAhomozygous iPSCs to maximize the coverage [164,165]. Another approach is to adopt genome-editing technologies to generate hiPSCs with enhanced immune compatibility to avoid immune rejection in allogeneic recipients caused by HLA mismatching. For example, a CRISPR/Cas9 system was recently used to delete HLA-A and HLA-B biallelically, while retaining a single haplotype of HLA-C and non-canonical HLA-E, -F, and -G of hiPSCs, the strategy of which was shown to be effective at suppressing CD8 + T and NK cell activities [166]. Alternatively, universal hiPSCs may be generated by the disruption of β2microglobulin (B2M) and class II transcription activator (CIITA) [167][168][169][170][171], causing major histocompatibility complex (MHC) class I and II inactivation, respectively, together with the introducing of HLA-E/B2M fusion or the CD47 transgene [169]. Notably, depletion of B2M alone would activate a response by NK cells as HLA class I molecules serve as an NK inhibitory signal. Collectively, these approaches provide hypoimmunogenic hiPSCs that can be further differentiated into various cell types without immunosuppression.
The potential applications of hiPSCs have also been demonstrated in combination with gene therapy for various hematologic disorders. The first proof-of-principle study was performed in a humanized, sickle cell anemia mouse model, where sickle cell manifestations can be rescued after transplantation of corrected hematopoietic progenitors derived from autologous iPSCs [172]. Further, the attempts to use hiPSC-based gene therapy have been demonstrated in various forms of SCID (JAK3-SCID, ADA-SCID, X1-SCID), in which corrected hiPSC-derived HSCs exhibited normal T cell development and could yield mature NK cells [118,119,173] and in CGD [110,112,115], as mentioned above. In addition, the feasibility of hiPSC-based gene and cell therapies has been demonstrated in hemophilia A, a bleeding disorder caused by mutations in the F8 gene. The researchers generated corrected hiPSCs using a lentiviral vector carrying the FVIII transgene and differentiated them into functional endothelial cell (ECs) followed by transplantation into immunodeficient mice. The hiPSC-derived ECs could produce functional FVIII and rescued the hemophilic phenotype after transplantation [127]. For hemophilia B, CRISPR/Cas9-mediated F9 gene correction was performed in patients' iPSCs prior to their differentiation into hepatocytelike cells (HLCs) and transplantation into a hemophilic mice model. The presence of FIX in corrected, transplanted hepatocytes was observed 6-9 months after transplantation [128].

Cell-Based Cancer Immunotherapy
Cancer immunotherapy is a transformative treatment for cancers that provides hope to patients with relapsed and refractory diseases, who are not responsive to conventional therapies. The concept of cancer immunotherapy arose from the tumor immune escape, one of the hallmarks of cancers, aiming to restore the host immune defense to keep cancer permanently at bay. Chimeric antigen receptors (CARs) are engineered transmembrane receptors that engage immune cells towards particular tumor-associated antigens. The clinical success stories of CAR-T cell therapy targeting B-cell malignancies-with four US FDA-approved autologous CAR-T cell products for the treatment of relapsed/refractory diffused large B-cell lymphoma (DLBCL), acute lymphoblastic leukemia (ALL), and mantle cell lymphoma-have made cellular immunotherapy an attractive field [174,175]. The number of CAR technology-related clinical trials has been increasing exponentially. Amidst this enthusiasm, several challenges to cellular immunotherapy from a technical perspective are: (i) the limited number of patients' immune cells obtained and their exhaustion; (ii) the difficulty in engineering primary cells and their fragility after the engineering process; (iii) the limited ability of primary immune cells to proliferate, which hinders clonal selection; and (iv) the personalized, autologous nature of certain immune cellular products to avoid the risk of deleterious GvHD [176]. Therefore, hiPSCs, which can easily be genetically modified and maintained practically indefinitely in culture and differentiated into any immune cells upon induction, are an ideal alternative cell source for developing next generation CAR-related immune cell therapy.
One important drawback of the current CAR-T cell therapy is its personalized, autologous form that is made for individual patients and cannot be prepared as off-the-shelf products. Hence, its noticeable limitation is associated with the manufacturing process, which is costly and laborious, increasing the risk of production failure in clinical settings, especially in those with a limited number of healthy T cells, rendering these CAR-T cells unsuitable for patients with rapidly progressing disease [177]. A previous study provided evidence that genetic engineering of hiPSCs with second-generation CARs would be an efficient strategy to generate functional and expandable CAR-T cells. To this end, hiPSCs were generated by reprogramming peripheral blood T lymphocytes from healthy donors and were subsequently transduced with second-generation anti-CD19 CAR and differentiated into T cells using the O9P9-DL cell lines. The differentiated anti-CD19 CAR-T cells displayed canonical features of T cell function and specificity towards CD19, suggesting the feasibility of generating CAR-T cells from hiPSCs [178]. Numerous efforts have been proposed to adopt genome-editing technologies to generate hiPSCs with enhanced immune compatibility, as mentioned above. Hence, subsequent incorporation of CARs into those hypoimmunogenic hiPSCs may enable the production of off-the-shelf, allogeneic CAR-T cells.
NK cells are part of the first line of defense and are rapidly activated without prior sensitization to protect the body against foreign materials and abnormal cells, including tumor cells. Allogeneic NK cells do not carry the risk of inducing GvHD, which is frequently associated with allogeneic CAR-T cells; therefore, they can be prepared as off-the-shelf products without further modifications [179]. Currently, several phase 1 clinical trials have utilized allogeneic hiPSCs as the source of NK cells for immunotherapy of various solid tumors and hematologic malignancies (NCT03841118; NCT04245722; NCT04106167; NCT04023071; NCT04551885). Of them, anti-CD19 CAR was expressed in hiPSCs together with a cleavable CD16 Fc receptor and an IL-15 receptor fusion prior to the induction of NK differentiation towards CAR-NK cells for the treatment of relapsed/refractory lymphoma and chronic lymphocytic leukemia (CLL) (NCT04245722) [180]. Notably, CAR-NK cells could eliminate tumors not only through the ability of CAR to target specific tumor antigens, but also through antigen-unrestricted killing activity of NK cells themselves, namely, granule-mediated cytotoxicity, cytokine-mediated cytotoxicity, and CD16-mediated antibody-dependent cell-mediated cytotoxicity (ADCC) [181].
In solid tumors, CAR-T cells have shown low success rates, mainly due to the high complexity of the tumor microenvironment containing an immunosuppressive network and physical and metabolic barriers [182][183][184]. Recently, immune cells of myeloid lineage, i.e., monocytes and macrophages, have gained increasing attention due to their potential capability of active accumulation in solid tumors and penetration into the dense stromal tissues surrounding tumors. It is noticeable that allogeneic CAR-macrophages carry a low risk of GvHD and can also be prepared off-the-shelf similar to those of CAR-NK cells [184]. Data from a phase 1 clinical trial of autologous anti-mesothelin CAR-macrophages for ovarian cancer and malignant peritoneal mesothelioma have shown preliminary safety with no treatment-related discontinuations (NCT03608618) [185], giving hope for the applications of novel cellular immunotherapy for solid tumors. The feasibility of the derivation of anti-CD19 CAR-macrophages-with antigen-dependent phagocytosis and antitumor functions in vitro and in vivo-from hiPSCs has recently been shown, although the efficacy and persistency need to be further improved by designing more suitable CARs [186].

Current Challenges and Future Perspectives
With the discovery of hiPSCs over the past decade, rapid progress has been made in biomedical and pharmaceutical research. hiPSCs open up research opportunities as a new platform for disease modeling and therapeutic drug development and screening; additionally, their translation to potential clinical applications, such as cell-based therapy and immunotherapy, has exponentially increased. The advantages of hiPSCs are even more pronounced when combined with other technologies, e.g., genome-editing technology, which helps either recapitulate or correct the disease phenotypes. However, several limitations remain to be further investigated.
For hiPSC-based disease modeling, patient-specific hiPSCs have been used as representative cells of patients that carry specific mutations, which could display particular disease phenotypes once differentiated in vitro. hiPSCs serve as an alternative to human primary cells that might not be available or easily accessible, and to complement animal models that might not resemble human pathophysiology. Thus far, numerous disease models using hiPSCs have been demonstrated; however, one of the concerns raised is the respective appropriate control-genetic variations by means of gene expression pattern and epigenetic status even in control hiPSCs were reported either from within or between individual donors [187,188]. These background genetic variations have a remarkable impact on the differentiation capacity of hiPSCs, resulting in a large heterogeneity of lineage phenotypes that confounded the interpretation of the data in disease modeling [189][190][191]. To this end, genome-editing technology, i.e., CRISPR/Cas9, was utilized to generate a genetically matched hiPSC line, known as isogenic correction control iPSCs, for use as comparative control cells in the study. Using this approach, variation in genetic background and other confounders would be eliminated. It is important to note that the possibility of potential off-target effects remains challenging. Alternatively, some studies have attempted to convert the primed state of hPSCs into a naïve state in order to overcome the heterogeneity, based on the findings in mouse ESCs. Several approaches have been used to induce a naïve state of hESCs, i.e., by adding a combination of kinase inhibitors, such as MEK, GSK3, ROCK, BRAF, and SRC in the presence of activin and hLIF [192] and overexpression of transcription factors NANOG and KLF2 [193]. However, these data are preliminary and concerns regarding the naïve state, e.g., genetic integrity and loss of imprinting, need to be further clarified.
Considering the ability of hiPSCs to differentiate into disease-relevant cell types and various cell products, it has been shown that certain differentiated cells, with most evidence from blood cells, exhibited immature functional characteristics. Several studies demonstrated that HSPCs differentiated from hiPSCs do not truly resemble bona fide HSCs with long-term repopulation, also known as definitive HSCs. In addition, terminal differentiation of iPSC-derived HSCs toward a specific lineage, such as an erythroid lineage, produced mostly erythroid cells with embryonic or fetal globins, and had a low enucleation rate. These limitations may slow the progress of hiPSC-based disease modeling, cell-based therapy, and drug discovery. To overcome these obstacles, high-efficiency differentiation protocols should be developed. It has been demonstrated that the hematopoietic niche is mainly involved in HSC production. A recent study showed that AGM-associated macrophages played roles in HSC production by providing the microenvironment for definitive EHT and hematopoietic maturation [191], suggesting that appropriate niches are very important for cellular development. A more recent study utilized single-cell transcriptomics to identify novel mechanisms that play a key role in definitive hematopoietic production, and cell cycle regulators, e.g., CDK4/6 and CDK1, have been identified as essential for EHT and hematopoietic differentiation [194]. Therefore, using a hiPSC differentiation system combined with other new technologies may provide more in-depth understanding in human hematopoiesis that could lead to the establishment of efficient differentiation methods/protocols for better production of functional blood cells in the near future. Of note, the use of small molecule inhibitors/enhancers is gaining interest as an attractive method for both reprogramming and improved cell-specific differentiation that are appropriate for clinical applications, due to the ease of use and zero risk of integrating exogenous genetic factors.
To date, the applications of hiPSCs for clinical purposes such as cell-based therapy are an active area of research. One of the critical issues is potential tumorigenicity, attributable to the residual undifferentiated and/or immature cells derived from hiPSCs existing in the final cell products, which could possibly lead to the emergence of teratomas or tumors after transplantation [195][196][197][198]. Many efforts have been made to overcome these challenges. For example, the establishment of efficient directed differentiation methods; quality control, particularly on the cell purification using positive and/or negative antibody selection; and the use of safety switches that selectively eliminate the tumorigenic cells. It has been shown that suicide genes, i.e., herpes simplex virus type 1 thymidine kinase (HSVtk), and drug-inducible safeguard systems could be used to deplete immature proliferating cells or undifferentiated hiPSCs and hiPSC-derived cells after transplantation [199,200], making hiPSC-derived cell therapy safer. Autologous hiPSC transplantation has been considered a promising tool for therapeutic application due to its safer profile of immune responses. However, the major drawbacks of this strategy are the cost of production and the limited availability/accessibility, which is critical for patients with rapidly progressing conditions, as death may occur prior to the cell therapy, e.g., from heart failure or spinal cord injury. These limitations could be potentially overcome by using allogeneic hiPSC transplantation, but only if the issues with immunogenicity have been resolved. Notably, comprehensive quality testing, including contamination tests, such as sterility and viral testing, morphology, HLA and STR analyses to prevent sample mix-ups, pluripotent markers, karyotyping, and genomic analyses to evaluate genomic mutations, are mandatory for clinical-grade hiPSCs.

Conclusions
In conclusion, hiPSC-based disease modeling, with cells either obtained from patientspecific somatic cells carrying specific mutations or from the introduction of normal iPSCs with specific mutations, offers an exclusive and convenient means to model disease phenotypes that also allows studies to track disease progression in the case of acquired genetic disorders. Although several concerns still limit the validity of hiPSC-based disease modeling, most of these limitations can be overcome by following some general guidelines or by mindfully interpreting the data. For example, the heterogeneity of hiPSCs and their differentiated progeny among different individuals can be overcome by using an isogenic, unaffected control whenever possible. Higher numbers of clones and better standardization of efficient differentiation protocols are necessary for reduced variability and increased reproducibility. Although the functional immaturity of the iPSC-derived blood cells remains an important issue to be addressed, hiPSC-based disease modeling provides a greater understanding of the pathophysiology, particularly early events that could potentially lead to the identification of novel therapeutic targets, and it is a useful tool for high-throughput screening of novel drugs. hiPSCs can also be viewed as a game changer for cell-based therapy. hiPSCs hold great promise for both autologous HSCs and iPSC-derived blood cells/products. However, an increasing trend has been the clinical applications of hypoimmunogenic, allogeneic iPSC-based cell therapy, attributable to the laborious production and time delay of personalized cells. Recently, CAR-T cell therapy has become a new hope for cancer treatment, leading to the rapid growth in cellular immunotherapy. hiPSCs have been listed as an ideal alternative cell source for cancer immunotherapy, especially when certain allogeneic immune cells, such as CAR-NK cells and CAR-macrophages, carry a low risk of immune rejection and GvHD. Overall, hiPSCs greatly benefit biomedical and biopharmaceutical research and offer great opportunities for future regenerative medicine.