Hematopoietic stem cells: Understanding the mechanisms to unleash the therapeutic potential of hematopoietic stem cell transplantation

Abstract

Hematopoietic stem cell transplantation (HSCT) is a promising approach in regenerative medicine and serves as a standard treatment for different malignant and non-malignant conditions. Despite its widespread applications, HSCT is associated with various complications that compromise patients’ lives and pose considerable risks of morbidity and mortality. Understanding the molecular physiology of HSCs is fundamental to ultimately enhance the mobilization, engraftment and differentiation of HSCs, thus unleashing the full therapeutic potential of HSCT in the treated patients. This review outlines the current understanding of HSC biology and its relevance to the clinical challenges associated with HSCT. Furthermore, we critically discuss the pros and cons of the preclinical murine models exploited in the HSCT field. Understanding the molecular physiology of HSCs will ultimately unleash the full therapeutic potential of HSCT. HSCs derived from induced pluripotent stem cells (iPSCs) might present an attractive tool which could be exploited preclinically and clinically. Nonetheless, further studies are warranted to systematically evaluate their potential in terms of improving the therapeutic outcome and minimizing the adverse effects of HSCT.

Introduction

Hematopoietic stem cell transplantation (HSCT) is a treatment option for various malignant and non-malignant diseases among children and adolescents. HSCT, also known as bone marrow transplantation, is an infusion of healthy hematopoietic stem cells (HSCs) in needy patients. The process involves collecting hematopoietic stem and progenitor cells (HSPCs) from bone marrow via direct bone marrow harvest or from peripheral blood via a process called apheresis [1]. The collection process is followed by processing HSPCs and transferring them back into patients. Transfused stem cells become engrafted when they make their way to the bone marrow and start producing new blood cells.

The first successful bone marrow transplant was performed in New York in 1957 between identical twins to treat acute leukemia [3]. A few years later, the first successful allogeneic bone marrow transplant was reported in Minnesota in 1968 to treat a 5-month-old child with severe combined immunodeficiency syndrome via engraftment from peripheral blood and bone marrow from a sibling donor [4]. Since then, HSCT has been used as standard therapy for various hematologic diseases worldwide. Peripheral blood stem cell (PBSC) transplantation has largely replaced BM transplantation in autologous and allogeneic cases (Fig. 1) [5]. HSCT is classified as allogeneic or autologous according to the donor status. In allogeneic transplantation, stem cells are collected from a donor other than the recipient, who is a human leukocyte antigen (HLA)-matched family member, an unrelated HLA-matched donor, or a mismatched family donor (haploidentical) [6]. Autologous bone marrow transplantation occurs when the stem cells are collected from the patient and re-infused after purification methods [7]. Syngeneic bone marrow transplantation occurs when the donor and recipient are identical twins [8, 9]. Umbilical cord blood transplantation is a third category of graft sources and involves stem cells that are collected from the umbilical cord immediately after the infant is born [10]. Compared with bone marrow-derived stem cells, umbilical cord-derived stem cells are associated with a lower risk of graft‐versus‐host‐disease (GVHD) [11]. However, because of the inadequate collected dose of HSCs from umbilical cord-derived cells, other transplantation settings are usually used.

Fig. 1.

Schematic representation of peripheral blood stem cell transplantation (PBSCT). PBSCT involves collection of peripheral blood stem cells from a healthy donor (allogeneic PBSCT) or a patient’s own cells (autologous PBSCT). In allogeneic collection, four to five days before stem collection, the donor begins receiving mobilization agents such as granulocyte colony-stimulating factor (G-CSF) that increase the release of stem cells from the bone marrow into the bloodstream for collection via the apheresis process, which separates the stem cells and returns the blood content back to the donor/patient [1, 12, 13]. In autologous collection, patients receive a combination of mobilization agents along with chemotherapy [1]. Before transplantation, the recipient undergoes a preparative regimen that involves high-dose chemotherapy and/or total body irradiation to eliminate diseased cells and induce immunosuppression to facilitate transplanted stem cell engraftment [14]. Finally, stem cells are infused into the patient's body. The schematic representation was created with BioRender.com

In this review, we critically discuss the clinical challenges associated with HSCT that hinder its full utilization. We also review the current understanding of HSCs’ self-renewal biology and its complexity. Further, this review critically sheds light on diverse preclinical models exploited to enhance our current comprehension of the molecular mechanisms underlying the survival, self-renewal, migration, and engraftment of functional HSCs to ultimately devise more efficacious and tolerable therapeutic strategies to unleash the full potential of HSCT.

Clinical applications of HSCT

HSCT improved the clinical outcomes of patients with severe diseases. HSCT, either autologous (using genetically edited patient stem cells) or allogeneic, is the only curative treatment for sickle cell anemia [15, 16]. Furthermore, autologous HSCT is a preferred curative option for multiple sclerosis (MS) because there is a lower risk of GVHD and infection [17]. Other common non-malignant indications for autologous HSCT are systemic sclerosis and sterile rheumatoid arthritis [18]. Autologous HSCT is the standard treatment for multiple myeloma (MM), and indeed, it increased the survival rate of MM patients [19, 20]. Non-Hodgkin lymphoma (NHL), Hodgkin lymphoma (HL), neuroblastoma, ovarian cancer, Ewing sarcoma, germ-cell tumors, osteosarcoma, Wilms tumor, and medulloblastoma are also malignant indications for autologous HSCT [18]. For allogeneic HSCT, the non-malignant indications include aplastic anemia, paroxysmal nocturnal hemoglobinuria, Fanconi anemia, blackfan–diamond anemia, thalassemia major, severe combined immunodeficiency, Wiskott–Aldrich syndrome, inborn errors of metabolism, metachromatic leukodystrophy, and primary immunodeficiency diseases [18]. The malignant indications for allogeneic HSCT are acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), chronic myeloid leukemia (CML), myelodysplastic syndrome, myeloproliferative disorders, NHL, HL, chronic lymphocytic leukemia (CLL), and MM [18].

Clinical challenges associated with HSCT

Despite its curative potential against various diseases, HSCT can also lead to several complications that may compromise patients’ lives and can cause death. Non-relapse mortality is driven by infection, GVHD, or organ failure [21, 22]. Various factors may increase the susceptibility to infection, such as granulocytopenia and compromised immune system; slow or impaired recovery of natural killer (NK), T, and B cells; mucositis; inflammation of mucous membranes that line the gastrointestinal tract (GI); events related to vascular access devices such as catheters; and organ dysfunction [23, 24].

Bacterial infections, including infections caused by gram-negative bacteria such as enteric bacteria and gram-positive cocci such as viridians and streptococci, are common infections in the first few weeks post-transplant [24]. The infection sites may include the bloodstream, lung, GI tract, and skin. Engraft recipients are at risk of viral infections such as cytomegalovirus, Epstein–Barr virus, or adenovirus [24]. Fungal infections, including candida, Aspergillus, other non-Aspergillus molds, and Pneumocystis jirovecii, can also occur in the period following transplants. Transplanted cells may also attack lung and renal cells, causing inflammation and organ failure [23, 24]. Dysbiosis is another HSCT-associated complication resulting from compromised GI mucosal barrier and altered microbiota due to the administration of different antibiotics.

GVHD is a common complication following allogeneic transplantation. Indeed, more than 10% of the patients die from GVHD complications [25, 26]. GVHD results from the recognition of the recipient body as foreign, causing immune attack by the transplanted cells. GVHD is classified as: i) acute GVHD, occurring within the first 100 days post-transplantation, or ii) chronic GVHD, presenting after 100 days of transplantation [27]. The common features of GVHD include apoptosis of the epithelial cells of the GI tract, small intestinal crypt dilation and distraction, neutrophilic infiltration, liver dysmorphology and inflammation, and separation of the dermis from the epidermis of the skin [28]. GVHD can also affect other organs, including the lungs, kidneys, and hematopoietic system. T cells are the primary players in GVHD, and current treatment options include donor T-cell immunosuppressors such as corticosteroids.

Currently, there are major gaps in our knowledge regarding the molecular mechanisms by which transplanted HSCs can induce these clinical complications and, thus, this restrain/compromise their optimal treatment. Indeed, further systematic studies are warranted to address this gap and objectively optimize the use of HSCT in clinical practice, reduce its adverse effects, and thus improve patient outcomes.

HSC quiescence and self-renewal biology

The hematopoietic system is maintained by the production of new blood cells arising from HSCs within the BM. HSCs have unlimited self-renewal and multipotency potential, making them suitable for transplantation into patients with defective blood cells [29]. Under physiological conditions, HSCs maintain a quiescent state, where cells undergo reversible cell cycle arrest and have low metabolic activity [30], to maintain the self-renewal potential, prevent cell exhaustion, and protect against DNA damage [31–34]. Indeed, this is a tightly regulated cellular state that maintains lifelong hematopoiesis and prevents HSC depletion. Quiescent HSCs have long-term self-renewal potential, termed long-term self-renewal HSCs (LT-HSCs) [35], and form a rare population of the cellular BM niche, representing approximately 0.01–0.04% of the hematopoietic BM population [36]. LT-HSCs localize adjacent to the membrane lining the inner surface of the bone, the endosteal region, where they are exposed to signaling from functional bone and marrow stromal progenitor cells (BMSPCs) and arterial vessels that provide a protective niche for the quiescent state of HSCs [37]. Upon their activation by hematological stress, such as blood loss, LT-HSCs undergo self-renewal to generate short-term HSCs (ST-HSCs), which have limited self-renewal potential and are committed to multilineage differentiation [38]. ST-HSCs further differentiate into hematopoietic progenitors that eventually generate mature effector cells, including platelets, erythrocytes, granulocytes, macrophages, dendritic cells, B cells, T cells, and NK cells [39]. Furthermore, the interaction with the extracellular matrix (ECM) within the endosteal area results in the formation of a more stabilized LT-HSCs pool that supports long-term self-renewal. The stiffness of the ECM decreases toward the sinusoid's peripheral area [40], which maintains active ST-HSCs [41]. The ECM within this area is rich in collagen IV, laminin, and fibronectin, which are secreted by sinusoidal endothelial cells (SECs) and adjacent mesenchymal stromal cells [42]. SECs express cell-surface molecules that support HSC migration from and into the bloodstream.

In vivo studies exploiting various genetically modified mice demonstrated several/distinct mechanisms regulating HSCs’ quiescent and self-renewal states. Heterozygous reduction in the expression of GATA, a zinc finger transcription factor expressed in HSCs and progenitors that regulates their development and differentiation, in mice increased the quiescent state of the stem-cell-enriched population, leading to a severe reduction in the production and expansion of HSCs [43, 44]. Pax transactivation domain-interacting protein (PTIP) is expressed in the hematopoietic system and maintains HSC quiescence by regulating lysosomal activity [45]. Indeed, functional exhaustion of HSCs by increasing their lysosomal degradative activity was observed in the Pax knockout mouse model, in which Ptip was specifically deleted in HSCs. Furthermore, PTIP cooperates with transforming growth factor-β signaling to promote HSC quiescence via its interaction with SMAD2/3 [45]. HSCs are null/devoid or express low levels of macrophage-1 antigen (MAC-1). Following HSCT in mice, a transient increase in MAC-1 expression was observed in HSCs isolated from murine BM. MAC-1 expression is inversely correlated with the cell cycle, and cells positive for MAC-1 are more quiescent than MAC-1-negative cells [46].

The activity of HSCs within the BM is controlled by various signaling pathways [54].The Wnt/β-catenin pathway is a core regulator of the self-renewal and differentiation of HSCs. When the Wnt protein binds to Wnt receptors, including frizzled (Fzd) and LRP5/6, downstream intracellular signaling pathways, which can be β-catenin dependent or β-catenin independent,are initiated. ِِِActivation of Wnt signaling induces the transcriptional regulation of target genes that regulate the stemness of HSCs [55]. Conversely, inhibition of Wnt signaling via PKF-115, a T-cell factor/β-catenin inhibitor, reduces the number of HSCs. Mg²⁺/Mn²⁺-dependent protein phosphatase 1B (Ppm1b) activates Wnt signaling by inducing the dephosphorylation of β-catenin [56]. Depletion of Ppm1b in murine hematopoietic cells led to disruptive quiescence and suppressed expansion which ultimately impaired HSC homeostasis and hematopoietic reconstitution [56]. Combined inactivation of Wnt signaling and Hoxa9 pathways resulted in severe defects in HSCs and their progenitors and impaired engraftment and reconstitution, suggesting that these two pathways work in a compensatory manner to regulate the HSC functional activity [34].

Cellular metabolism is a crucial process that regulates HSC biology and has recently attracted much attention. The metabolic activity of HSCs is critical for their transition from quiescence to activation [57]. High mitochondrial NADPH levels, resulting from fatty acid oxidation (FAO) within the mitochondrial matrix, maintain cholesterol synthesis, which drives extracellular vesicle (EV) biogenesis [58, 59]. EVs support the self-renewal potential of HSCs in an autocrine manner. In addition, the transcription factor Nynrin is highly expressed in HSCs and is a critical regulator of HSC quiescence. Nynrin-deficient HSCs exhibit increased mitochondrial dysfunction and compromised self-renewal capacity [60]. Mechanistically, Nynrin transcriptionally represses the expression of the Ppif gene, which encodes for cyclophilin D, an essential component of the mitochondrial permeability transition pore (mPTP) that mediates HSC fate decisions [57, 60–63]. Thereby, Nynrin-mediated inhibition of PPIF inhibits mPTP opening, ROS production, and subsequent cell necrosis, thus preserving HSC frequency and maintaining their quiescence.

Despite considerable research efforts, the precise mechanisms by which HSCs maintain a balance between quiescent and active states are still not fully elucidated. A deeper understanding of HSC biology and functional aspects would ultimately improve the therapeutic outcomes of HSCT.

Preclinical murine models of HSCT

Mouse models are indispensable tools that play fundamental roles in the inception, development, and optimization of HSCT. The basic protocol of experimental HSCT comprises the transplantation of human/mouse-derived HSCs into pre-conditioned recipient mice (Fig. 2A-2D). Immunocompetent mice are usually pre-conditioned with total body irradiation before their transplantation with murine donor HSCs,whereas immune-deficient mice are either sublethally irradiated or not irradiated prior to their transplantation with human donor HSCs (Fig. 2A-2D). The engrafted HSCs are then tracked/monitored at different time points in the peripheral blood, BM, and/or the whole body of the recipient mice to evaluate the contribution and function of the transplanted HSCs. Murine models of BMT and GVHD have been extensively reviewed elsewhere [64–66]. In this section, we focus on the lessons learnt from the preclinical mouse models of HSCT.

Fig. 2.

Schematic representation of the preclinical models of hematopoietic stem cell transplantation (HSCT). A In congenic HSCT, hematopoietic stem cells (HSCs) are collected from the bone marrow (BM) of the femur and/or the tibia of donor C57BL/6 CD45.2 mice. HSCs are then intravenously injected into recipient C57BL/6 CD45.1 congenic mice, which are preconditioned by irradiation or chemotherapy. B In allogeneic HSCT, HSCs are collected from the femur and/or the tibia of donor C57BL/6 mice. HSCs are then intravenously injected into recipient BALB/c mice, which are preconditioned by irradiation or chemotherapy. C–D In xenogeneic HSCT, human HSCs are obtained from the cord blood, the mobilized peripheral blood, or the bone marrow of a donor human patient. HSCs are then intravenously injected into recipient immune-deficient mice (C: irradiated NSG mice and D: NBSGW mice)

Murine HSCT models

In these models, murine HSCs are engrafted into syngeneic, congenic, or allogeneic recipient mice (Fig. 2A-2 B) [67]. Taking advantage of the allelic variants of the CD45 markers (namely, CD45.1 and CD45.2), which are expressed on all hematopoietic cells with the exception of mature erythrocytes and platelets, the C57BL/6 (CD45.1⁻ CD45.2⁺), BoyJ (CD45.1⁺ CD45.2⁻), and F1-crossed hybrid C57BL/6 × BoyJ (CD45.1⁺ CD45.2⁺) mouse strains are exploited to examine the contribution/engraftment of donor, competitor, and recipient cells [64–66, 68, 69].

Exploiting CD45.1/2-co-expressing mice revealed a preferential bias toward CD45.2 reconstitution, which was not linked to an immunogenic response to the CD45.1 epitope [68]. Notably, this study pinpointed sex bias, with female mice displaying more CD45.2 reconstitution than male recipient mice do [68]. Murine HSCT models demonstrated that the peptidase CD26 (dipeptidylpeptidase IV) abrogated the homing and engraftment of HSCs [70]. Indeed, genetic depletion or pharmacological inhibition of CD26 (using diprotin A, a specific CD26 inhibitor) augmented the homing of CD45.2 BM-derived HSCs to CD45.1 recipient mice [70]. Increasing α1,3-fucosylation increased the repopulating capacity of murine BM-derived CD45.2 CD34⁺ HSCs in CD45.1 C57BL/6 recipient mice [71].

Cystinosis is a metabolic disorder characterized by the accumulation of cystine secondary to a defective CTNS gene, which encodes the lysosomal cystine transporter (cystinosin) [72]. HSCT of a murine model of cystinosis (irradiated C57BL/6 Ctns⁻/⁻ mice) with BM-derived HSCs (obtained from green fluorescent protein (GFP)-transgenic C57BL/6 mice) significantly reduced cystine accumulation [72]. Transplantation of C57BL/6 mice with genetically modified HSCs stably expressing Nogo receptor(1–310)-Fc ameliorated the neurological decline and histopathological anomalies observed in experimental autoimmune encephalomyelitis [73].

Allogeneic HSCT-associated acute renal damage was modeled in vivo by exploiting the disparity in MHC class I and II antigens between BALB/c (H-2^d) donor mice and C57BL/6 (H-2^b) recipient mice [74]. Another GVHD model was established by transplanting splenocytes from C57BL/6 (H-2^b) donors into B6D2F1 (H-2^b/d) recipient mice [74]. Minnelide (a prodrug of triptolide with antitumor and anti-inflammatory capabilities) ameliorated GVHD in a complete MHC-mismatched (C57BL/6 → BALB/c) allogeneic HSCT model and in MHC-matched, minor antigen–disparate donors/recipients (C3H.SW → C57BL/6) model of GVHD [75]. Diet-induced obesity had deleterious effects on the gut microbiota which may affect GVHD after allogeneic HSCT in murine models. Notably, these effects were attenuated using prophylactic antibiotics [76]. Adoptive transfer of a large number of donor regulatory T cells (Tregs) conferred superior efficacy compared to post-transplant cyclophosphamide in terms of ameliorating acute GVHD following allogeneic HSCT in vivo [77]. Zinc supplementation and activation of the zinc-sensing receptor GPR39 promoted the thymic function and T-cell reconstitution after allogeneic HSCT in C57BL/6 and Balb/c mice [78].

Internal tandem duplication mutation in the FLT3 tyrosine kinase receptor (FLT3-ITD) is one of the most common mutations in AML [79, 80]. Pre-transplantation of allogeneic BM and spleen grafts with anti-human CD4 MAX.16H5 IgG1 prior to their co-transplantation with 32D-FLT3-ITD AML cells abrogated GVHD and prolonged the survival of treated C3H/HeN mice [81]. Similarly, pre-treatment with anti-human CD4 MAX.16H5 prevented allogeneic GVHD in P815-BALB/c leukemic mice without compromising the graft-versus-leukemia effect of BM and splenocytes obtained from CD4^−/− C57BL/6 mice transgenic for human CD4 and HLA-DR3 [82]. Combo-regimens of anti-CD45 radioimmunotherapy using an α-emitter (such as astatine-211 [²¹¹At]) together with syngeneic BMT prolonged the survival of mice with disseminated leukemia [83]. Targeting interleukin-2-inducible T-cell kinase maintained graft-versus-leukemia efficacy against primary B-cell acute lymphoblastic leukemia but abrogated GVHD during allogeneic HSCT in vivo [84].

The advantages of murine HSCT models include their low cost and provision of in vivo settings/platform recapitulating syngeneic, congenic or allogeneic HSCT scenarios [85]. Nonetheless, interspecies differences and lack of human targets compromise the clinical relevance of these models [64–66].

Humanized (or xenogeneic) HSCT models

Given the limitations of murine HSCT models, considerable efforts have been invested in devising humanized models, mostly through engrafting human HSCs into immunodeficient mice (Fig. 2C-2D) [85]. Compared with adult BM-derived HPSCs, human CB cells are considered the most commonly exploited source of HPSCs owing to their relative availability and favorable engraftment rates [85]. For example, severe combined immunodeficiency (SCID) mice which harbor a mutation in the gene encoding for protein kinase, DNA-activated, catalytic polypeptide (PRKDC), the catalytic subunit of DNA-dependent kinase (DNA PKcs) in the non-homologous end-joining (NHEJ) pathway [86])—lack functional B and T cells. SCID-based models have been employed extensively to hasten HSC engraftment. Indeed, the first engraftment of human CD34⁺ cells was conducted in SCID mice. However, it resulted in low engraftment rate, lack of human cytokine production, and limited HSC differentiation capability that is restricted to T and B cells.

An improved immunodeficiency mouse model was then generated by crossing the scid mutation into a non-obese mouse (NOD/SCID) model. This model exhibited decreased NK cell and innate immune functions that ultimately promoted the engraftment and multilineage differentiation of human HSCs without the need for exogenous administration of cytokines. However, this model has several limitations, including the need for pre-conditioning with irradiation, partial generation of the human immune system, and limited development of human T cells.

NOD-SCID IL2rγ^null (NSG) mice were generated by inducing mutations in the IL2-receptor common γ chain (IL2rg) gene in NOD-SCID mice. This enhanced immunotolerance to engrafted HSCs and improved multilineage differentiation. The functions of several interleukin receptors, such as IL2, IL4, IL7, IL9, IL15, and IL21, require the IL2rg chain for ligand binding [87]. Minnelide decreased the donor T-cell frequency within the peripheral lymphoid compartment following xenogeneic human-to-mouse transplantation (mobilized human PBMCs → irradiated NSG mice) [75]. Fourteen days after transplanting NSG mice with human PBMCs, Sligar and colleagues injected THP-1 AML cells and tested the potential effect of post-transplant cyclophosphamide when administered alone and together with tocilizumab (an anti-IL6) [88]. Indeed, post-transplant cyclophosphamide alone or together with tocilizumab prolonged the survival of hPBMC leukemic transplanted mice indicating their capability to impair GVHD without abrogating the therapeutic effects of the graft-versus-leukemia.

Wang and colleagues developed the so-called HuNSG model in which they engrafted NSG mice with human HSPCs, which gave rise to diverse human hematopoietic and immune cells and exhibited a partially functional human immune system [89].

Six to eight weeks following the engraftment of irradiated NSG mice with human CD34⁺ HSCs, patient-derived triple-negative breast cancer (TNBC) biopsies were transplanted into the mammary fat pads of humanized NSG mice [90]. The slightly slower growth rate of TNBC engrafted in humanized NSGs compared to non-humanized NSGs was speculated to be due to the acquisition of immunocompetency and the partially matched HLA typing between the transplanted patient-derived TNBC and human CD34⁺ HSCs [90].

Among the shortcomings associated with the use of these models is the lack of critical growth factor receptors and cytokines needed for the maturation and differentiation of myeloid lineages. To this end, a triple transgenic NSG-SGM3 mouse model, which constitutively expresses human granulocyte–macrophage colony-stimulating factor (GM-CSF), interleukin 3 (IL-3), and stem cell factor (SCF), was developed [91]. By crossing the NSG strain with C57BL/6 in the Kit^W41 strain, McIntosh and colleagues developed the NBSGW strain, which was engrafted with human CB CD34⁺ cells without irradiation [92]. Compared with the NSG strain, superior engraftment of human HSCs (CD34⁺CD38^low and CD34⁺CD90⁺) was reported in the BM of NBSGW mice [93]. Consistently, NBSGW mice, but not NSG mice, were more readily engrafted with human CB-derived CD34⁻ HSCs [94]. Fostering α1,3 fucosylation augmented the engraftment of human CB-derived CD34⁻HSCs in NBSGW mice but not NSG mice [94].CRISPR mediated gene correction of beta-globin in patient-derived HSCs together with autologous HSCT ameliorated erythropoiesis in a humanized model of sickle cell disease [95].

In conclusion, although murine and humanized mouse models significantly enriched our present comprehension of HSCs biology and HSCT, these models are still in their infancy stage, and further optimizations are warranted to exploit their full potential. Indeed, diverse protocols are proposed to optimize the in vivo preclinical testing of HSCs as well as the expansion of HSCs for clinical applications [96]. Wilkinson and colleagues developed an albumin-free culture system that comprises polyvinyl alcohol, high levels of thrombopoietin, and low levels of SCF and fibronectin [97]. Notably, this culture system supported the long-term ex-vivo engraftment and expansion of functional mouse HSCs in recipient mice which were not pre-conditioned. Humanized ossicles (or bone marrow‐like organoids derived from primary human mesenchymal stem cells (MSCs)) were embedded in scaffolds and then transplanted into mice [98]. These humanized ossicles were used to: i) investigate HSCs-niche interactions within physiological and pathological settings and ii) promote HSCs expansion. Zhou and colleagues developed a 3-dimensional (3D) engineered scaffold for co-culturing CD34⁺ HSPCs in vitro [99].

HSCT conditioning regimens also merit further optimization. Indeed, myeloablative irradiation distorts the BM microenvironment and triggers inflammatory signaling, which ultimately compromises the engraftment, self-renewal, and differentiation of HSCs [96]. Sublethal irradiation deleteriously affects the vascular and perivascular cells of the BM. Likewise, myelosuppressive drugs provoke inflammatory signaling and activate osteoclasts [100].Omer-Javed and colleagues demonstrated that mobilizing endogenous HSPCs from their BM niche could favor the engraftment of gene-edited HSCs without pre-conditioning the mice with chemotherapy [101]. Co-culturing of HSCs with BM-derived stromal cells was also proposed to abrogate the effects of myeloablative conditioning on the BM niche. A CD45-targeted antibody–drug conjugate, when administered as a monotherapy, conferred the benefit of myeloablative conditioning in murine models of allogeneic HSCT [102]. NBSGW mice were also engrafted with human CB CD34⁺ cells without the need of prior irradiation [92].

Pluripotent stem cell models for HSCs

Previous studies relied mainly on murine models which, despite their value, do not mimic/fully recapitulate the human hematopoietic system. In fact, much of our understanding of the hematopoietic system was derived from studies conducted in mouse models. However, humans and mice have significant genetic differences, limiting the applicability of studying human-derived grafts in mice. Humanized mouse models partially mimicked the human immune system. However, key challenges, such as cross-reactivity between mice and humans and the compromised potential of transplanted human HSCs, still limit their potential in preclinical research in terms of systematically dissecting the complex mechanisms underlying the development, activation, differentiation, adhesion, and migration of human HSCs. Therefore, further optimization of these humanized models is urged to develop more clinically relevant systems capable of systematically dissecting the molecular basis underlying the quiescent, activation, and dynamics of HSCs.

During the past decades, pluripotent stem cells (PSCs) gained remarkable attention, given their promising applications in regenerative medicine and preclinical research studying various human diseases, including hematological diseases. Indeed, induced pluripotent stem cells (iPSCs) can be effectively employed as future therapeutic options to overcome the clinical limitations associated with HSCT. iPSCs are pluripotent stem cells that are reprogrammed from somatic cells via enforced expression of stem cell transcription factors [103]. iPSCs have self-renewal potential and can differentiate into three germ layers: ectoderm, mesoderm, and endodermal lineages. Notably, several research groups developed distinct protocols that derived HSCs and blood cells from iPSCs. Using a combined approach of morphogen-driven differentiation and transcription factor-mediated cell fate conversion, Sugimura and colleagues were able to differentiate PSC to HSPCs [104]. They reported that seven transcription factors governed the endothelial-to-hematopoietic transition (EHT). Additionally, overexpression of a single transcription factor, MLL-AF4, in iPSCs promoted the development of long-term engraftable HSPCs with multilineage reconstitution ability [105]. Nonetheless, given the oncogenic role of MLL-AF4 in AML, special caution should be exercised while analyzing the results stemming from these human HSPCs. Several other protocols have been reported to drive the differentiation of iPSCs into different immune cells. For example, via the bead-based approach, T-lineage cells were derived from iPSCs via the induction of mesoderm and hematopoietic specification in iPSC-derived embryonic bodies that were incubated with DLL4 Fc-fusion protein (DL4-Fc)-coated microbeads. DL4-Fc induced Notch activation, which initiated the T-cell program [106]. DL4 is normally expressed in thymic epithelial cells and induces T-cell differentiation via interaction with the Notch-1 receptor on progenitor cells [107]. Immune cells derived from PSCs, including iPSCs and embryonic stem cells (ESCs), may provide clinical-grade cells that are homogenous and can be genetically modified. This can overcome the clinical limitations associated with patient-derived primary immune cells, such as NK and T cells, which represent a heterogeneous population with donor variability. In principle, HSPCs derived from iPSCs could be used to perform drug screening assays to identify hits (lead compounds or small molecules) that modulate the self-renewal, adhesion, migration and/or function of HSCs to ultimately enhance HSCT. Genetically modified HSCs derived from iPSCs could also be transplanted into humanized mouse models. Nevertheless, further studies are warranted to systematically devise/define optimal standardized protocols which can be used for preclinical and clinical applications.

Conclusion

Although HSCT offers a curative option for many diseases, post-transplant adverse events remain critical challenge. Understanding the mechanisms that regulate the activities of HSCs is crucial for limiting the adverse effects and improving the clinical outcomes of HSCT. Despite serving as a great tool, preclinical mouse models still need further optimization to better mimic the human hematopoietic system and reflect different clinical scenarios of HSCT. hiPSCs could be exploited to systematically study HSCT within a more clinically relevant (i.e. patient-derived) context. Nonetheless, further studies are needed to improve the efficiency of generating clinical-grade HSCs from iPSCs which would ultimately pave the road for their transition from the bench to the patient’s bedside.

Abbreviations

AML

Acute myeloid leukemia

BM

Bone marrow

GVHD

Graft versus host disease

HLA

Human leukocyte antigen

HSCs

Hematopoietic stem cells

HSCT

Hematopoietic stem cell transplantation

HSPC

Hematopoietic stem and progenitor cells

PBSC

Peripheral blood stem cell

G-CSF

Granulocyte colony-stimulating factor

MSCs

Mesenchymal stem cells

MS

Multiple sclerosis

NK

Natural killer cells

hESCs

Human embryonic stem cells

hiPSCs

Human induced pluripotent stem cells

EBs

Embryonic bodies

DL4-Fc

DLL4 Fc-fusion proteins

GI

Gastrointestinal tract

LT-HSCs

Long-term self-renewal HSCs

BMSPCs

Bone and marrow stromal progenitor cells

ST-HSCs

Short-term HSCs

ECM

Extracellular matrix

SECs

Sinusoidal endothelial cells

PTIP

Pax transactivation domain-interacting protein

MAC-1

Macrophage-1 antigen

Fzd

Frizzled

FAO

Fatty acid oxidation

EV

Extracellular vesicle

mPTP

Mitochondrial permeability transition pore

TNBC

Triple-negative breast cancer

EHT

Endothelial-to-hematopoietic transition

Author contributions

AAA: Conception, literature review, writing–original draft, writing–review and editing. AKA: literature review, writing–original draft, writing–review and editing.

Funding

Not applicable.

Availability of data and materials

Not applicable.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.