Differentiation of human induced pluripotent stem cells into erythroid cells

During the last years, several strategies have been made to obtain mature erythrocytes or red blood cells (RBC) from the bone marrow or umbilical cord blood (UCB). However, UCB-derived hematopoietic stem cells (HSC) are a limited source and in vitro large-scale expansion of RBC from HSC remains problematic. One promising alternative can be human pluripotent stem cells (PSCs) that provide an unlimited source of cells. Human PSCs, including embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs), are self-renewing progenitors that can be differentiated to lineages of ectoderm, mesoderm, and endoderm. Several previous studies have revealed that human ESCs can differentiate into functional oxygen-carrying erythrocytes; however, the ex vivo expansion of human ESC-derived RBC is subjected to ethical concerns. Human iPSCs can be a suitable therapeutic choice for the in vitro/ex vivo manufacture of RBCs. Reprogramming of human somatic cells through the ectopic expression of the transcription factors (OCT4, SOX2, KLF4, c-MYC, LIN28, and NANOG) has provided a new avenue for disease modeling and regenerative medicine. Various techniques have been developed to generate enucleated RBCs from human iPSCs. The in vitro production of human iPSC-derived RBCs can be an alternative treatment option for patients with blood disorders. In this review, we focused on the generation of human iPSC-derived erythrocytes to present an overview of the current status and applications of this field.


Introduction
Blood transfusion is the main therapeutic option and a crucial part of modern medicine for patients with severe anemia [1,2]. A limited resource of blood, blood group compatibility (ABO and Rh antigens), and the risks of infection can present great challenges for blood transfusion [2,3]. Therefore, any alternative solution methods would be most helpful for patients with rare blood groups [4]. Mature red blood cells (RBCs) or erythrocytes/erythroid cells in a complex process called erythropoiesis are produced from hematopoietic stem cells (HSCs) [5][6][7]. Erythroblasts (precursors of RBCs) are difficult to proliferate in vitro [2,8,9]. In past decades, several groups have generated erythrocytes from umbilical cord blood (UCB)-derived HSCs [2,10]. Although multipotent HSCs have the capacity for self-renewal, the large-scale in vitro/ex vivo HSCs expansion and differentiation into RBCs is a difficult task [2,11,12]. Ex vivo cultured RBCs can also be obtained from immortalized erythroid precursors and pluripotent stem cells (PSCs) [13,14]. Human PSCs including embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) have the potential to proliferate indefinitely in culture and give rise to lineages of ectoderm, mesoderm, and endoderm [15][16][17]. Therefore, much attention has focused on human PSCs to replace current transfusion banking [18,19]. Several previous studies have revealed that human ESCs can differentiate into functional oxygen-carrying erythrocytes with normal function [18,20,21]. Unfortunately, the ex vivo expansion of human ESC-derived RBC is ethically and politically controversial [22,23]. In contrast, human iPSCs have less ethical and social issues compared to human ESCs [24,25]. Human iPSCs are produced by the manipulation of differentiated somatic cells [26][27][28][29]. Reprogramming of human somatic cells through the ectopic expression of transcription factors has provided a new avenue for disease modeling and regenerative medicine [16,30]. As human iPSCs have similar properties with human ESCs, these cells can be a suitable therapeutic choice for the in vitro/ex vivo manufacture of RBCs to eliminate blood supply shortages [31,32]. Various techniques have been developed to generate enucleated RBCs from human iPSCs [31,33,34]. Genome editing and human iPSCs technology has greatly accelerated the use of autologous transfusion therapies [35][36][37][38]. In this review, we focused on the generation of human iPSC-derived erythrocytes to present an overview of the current status and applications of this field.
Up to date, several culture systems have been established to obtain a sufficient number of mature and functional RBCs in vitro [8]. Three culture systems including erythroid cells lines (the murine erythroleukemia cell line and the human K562 cell line), HSCs derived from peripheral blood (PB) and UCB, and stem cells (human ESCs, neonatal cord blood (CB), mesenchymal stem cells (MSCs), and human iPSCs) have been evaluated to obtain RBCs [26,45]. An immortalized or continuous cell lines have a homogenous karyotype that can be altered following continuous cell culture, which might not be the most appropriate for pre-clinical studies or clinical applications [46]. In contrast, human PSCsderived RBCs can be achieved in larger scale cultures [47,48]. Many attempts have been made previously to establish all blood lineages, including lymphocytes, megakaryocytes, neutrophils, and RBCs from human ESCs [49][50][51][52]. Kaufman et al. for the first time revealed that human ESCs on the murine bone marrow cell line or a yolk sac endothelial cell line could give rise to erythroid cells [53]. Similar results were investigated by other lab using human fetal liver cells to improve the yield of human ESCs-derived erythroid cells [54]. However, the use of human ESCs has faced several challenges, including the host immune response toward ESCs and the ethical issues associated with the destruction of human embryos [55]. Differentiation of mouse iPSCs to blood cells has been investigated less intensively compared to the ESCs [56,57]. Recently, human iPSCs have been studied as one of the potential sources for HSCs and RBCs [58][59][60].
Although HSCs can be used for the treatment of hematological disorders [66,67], the bone marrow biopsy is an invasive procedure with chronic graftversus-host diseases (GVHD), morbidity, and mortality in patients who received allogeneic HSC therapy [68,69]. Autologous HSCs are an alternative option with a lower mortality rate, though in some cases, genetic correction is necessary before autologous HSCs transplantation [70][71][72]. However, in vitro expansion of HSCs is one of the main hurdles of autologous HSCs [73,74]. These limitations can be solved with a renewable source of cells. Human iPSCs as unlimited supplies of autologous cells can be an ideal candidate for genetic correction, differentiation to healthy HSCs, and autologous transplantation [67,75].
More recently, established iPSCs from human fibroblast cells represent a powerful tool for the investigation of early hematopoiesis [59,76,77]. One of the promising strategies for the use of iPSC is their capacity to differentiate into RBCs and to eliminate the allogeneic blood shortages [78,79]. Two approaches including embryoid body (EB) formation (feeder-free culture) and co-culture of human iPSCs with feeder cells have been employed for the generation of HSCs from human iPSCs. Many studies have aimed to differentiate human iPSCs into RBCs using EB formation [31,33,34]. In general, there are three steps for differentiation of iPSCs into RBCs, including the generation of iPSCs, HSCs, and mature RBCs [33,80] (Fig. 2). Many attempts have been made previously to achieve human iPSC-derived RBCs under conventional culture methods with SCF, EPO, VEGF, insulin-like growth factor I (IGF-1), dexamethasone (glucocorticoid receptor agonist), ITS (insulin, transferrin, and selenium), TPO, FLT3, BMP4, IL-3, IL-6, and EPO (Table 1). However, an ideal culture condition for human iPSC-derived RBCs should be able to generate large numbers of functional enucleated erythrocytes [31,86]. Feeder cells as a major cellular component have been found to enhance hematopoiesis from human iPSCs [81,82]. It has been shown that OP9 feeder cells as a mouse bone marrow stromal cell line may enhance the hematopoietic differentiation of human iPSCs [86]. Also, C3H10T1/2 feeder cells have the capacity to stimulate the hematopoietic differentiation of human iPSCs [38,87]. Increasing in vitro evidence indicates that the cell type of origin and an epigenetic memory for iPSCs may influence on the hematopoietic differentiation of human iPSCs [34,88,89]. Compared with iPSC-derived fibroblast cells, the human CD34+ hematopoietic population with the features of MSCs might be more suitable for the hematopoietic differentiation of iPSCs [81]. Following differentiation, HSCs lose their repopulation capacity. Hence, CD34+ HPCs must be purified before starting the differentiation [38]. Human iPSC-derived CD43+ hematopoietic cells have a strongly glycosylated transmembrane sialomucin that can be a suitable option for in vitro erythropoiesis [79]. In HSCs, reactive oxygen species (ROS) can modulate a balance between proliferation and differentiation. In the early stage of hematopoietic differentiation, mitochondria and NADPH oxidases (NOX) are the main sources of ROS [90,91]. NOX4 as the major NOX enzyme have been shown to play a significant role in the early stages of hematopoietic differentiation from iPSCs [85]. UM171 is a potent small molecule (HSC self-renewal agonist) that increases the derivation of HSPCs from human iPSCs in vitro [84,92]. Choi    illustrated the large-scale expansion of human iPSCderived erythroid cells under feeder-free and serum-free culture conditions [83]. They used several small molecules such as StemRegenin (SR1, a dual RasGAP and ERK1/2 inhibitor), BIO (archetypal GSK3b inhibitor), CHIR99021 (GSK3b inhibitor), IBMX (nonspecific inhibitor of cAMP and cGMP phosphodiesterases), and A-A014418 (GSK3b inhibitor VIII) to promote erythroid differentiation of human iPSCs [83]. Recently, Bernecker et al. described a simplified cell culture system with low cytokine support (SCF, EPO, and IL-3) to generate prolonged human iPSCderived RBCs [31]. Tursky et al. compared four serum and feeder-free iPSC hematopoietic differentiation protocols and investigated that two-dimensional (2D)-multistep protocol was simple and time-and cost-effective with the most efficient CD34+ progenitor cells [93].

Primary technical challenges for the clinical application of iPSC-derived RBCs
The in vitro production of human iPSC-derived RBCs can be an alternative treatment option for patients with blood disorders [94]. Many attempts have been examined to differentiate iPSCs into RBCs, but no clinical trials using iPSC-derived RBCs transfusion have been conducted [60]. Table 2 shows patient-specific iPSCs models of hematological disorders.
Before iPSC-derived RBCs derivatives can be used in the clinic, it is essential to found the risks and processrelated challenges associated with the generation of latestage maturity RBCs in vitro [132,133]. The technology of manufacturing functional erythroid cells from iPSCs needs a sufficient number of functional RBCs in a serum free-liquid culture system or chemically defined media, which is necessary for any potential clinical trials [134,135]. Human iPSCs may be considered as an unlimited source of RBCs than HSCs, but generating mature RBCs from iPSCs is still an inefficient process and less strict experiment protocols using low-cost media and reagents are needed [136]. Thus, the challenge for large-scale expansion of iPSC-derived erythroid cells needs to be overcome [8,60].
The use of small molecules as substitutes for growth factors or various cytokines can reduce side effects and media costs [137,138]. Further studies are necessary to understand which genetic or epigenetic alternations improve the terminal differentiation of iPSC-derived erythroid cells [139]. Recent studies have shown that histone deacetylases such as histone deacetylase 2 (HDAC2) are the critical regulator for chromatin condensation in mouse erythroblasts [140]. Administration of HDAC2 inhibitors can suppress the terminal differentiation of human erythroid precursors [139]. It is    therefore conceivable that HDAC2 activators may enhance chromatin condensation of iPSC-derived erythroid cells [141]. MicroRNAs are important regulators that downregulate the expression of target genes [142,143] and improve the maintenance of immature hematopoietic cells and terminal erythroid differentiation [58,143]. Therefore, different combinations of microRNAs may increase the numbers of iPSC-derived mature RBCs [139]. In addition to microRNAs, long noncoding RNAs (lncRNAs) have recently been reported that can determine the fate of stem cells [144]. A recent study has shown that long intergenic noncoding RNA erythroid prosurvival (lincRNA-EPS) can suppress apoptosis and facilitate erythropoiesis [145][146][147]. In this regard, lncRNAs might allow the generation of functional and mature RBCs from iPSCs [148,149]. Several groups have recently shown that 3D scaffolds such as poly (D, L-lactide-co-glycolide), polyurethane, collagen type I, and porous polyvinyl fluoride resin can mimic the bone marrow niche and improve maintenance of immature hematopoietic cells [150][151][152][153][154]. Although the in vitro maturation of iPSC-derived RBCs still presents several barriers, the cultured erythroid cells from iPSCs provide an important step toward fully defined and animal-free cultivation protocols that can be applied for transfusion medicine [67]. Figure 3 shows new technologies toward the large-scale expansion of human iPSC-derived erythroid cells.

Conclusion
Reprogramming of somatic cells to the pluripotent state has been suggested as an alternative source and a novel opportunity for patient-specific stem cell-based therapies, modeling of RBCs diseases, and drug testing [155]. Previous studies have shown that human iPSCs can give rise to erythroid cells, while in vitro derivation and maintenance of enucleated erythrocytes have still been challenging [86]. Also, many hurdles such as reprogramming without retroviruses, large scale and cost-effective production of iPSC-derived enucleated RBCs, and defined xenogenic-free conditions remain to be improved before human iPSC-based therapy [156,157]. According to good manufacturing practice (GMP) guidelines, the establishment of iPSC-derived RBCs using a reproducible, defined, and simple method will ease the translation of iPSCs into the clinic [93,158,159].

Funding
Not applicable.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Competing interests
The authors declare no conflict of interest.