Hemophilia A (OMIM #306700) is a monogenic, X-linked recessive bleeding disorder induced by the deficiency or reduced activity of FVIII, with a prevalence of 1:5000 amongst the male population (Graw et al., 2005). The most frequent genetic change that affects the F8 gene, located at Xq28, is the intron 22 Inversion (Inv22), accounting for approximately 45% of patients with severe Hemophilia A (Oldenburg and El-Maarri, 2006). In cases of mild impairment, the residual FVIII activity ranges from 5% to 40%, and patients experience intense bleeding upon injuries, traumas, and surgeries. The hemorrhagic episodes of moderate and severe cases, correlating to 1%–5% and <1% FVIII activity, respectively, include spontaneous and recurrent bleedings as a result of minor endothelial injuries (White et al., 2001). The current Hemophilia A therapeutic treatment mainly consists of protein replacement, requiring frequent prophylactic or on-demand parenteral administrations of recombinant FVIII (r-FVIII) with an extended half-life or plasma-derived FVIII (Powell et al., 2012). However, neutralizing antibodies against FVIII develop in nearly 30% of the cases, leaving these patients with very limited therapeutic alternatives. Interestingly, Hemophilia B (OMIM # 306900) is clinically analogous to Hemophilia A, but correlates to deficient or dysfunctional FIX (Jaloma-Cruz et al., 2012). It has a prevalence of 1:25,000 males, and it is caused by mutations in the F9 gene, at Xq27.1-27.2. The molecular mechanisms are heterogeneous and include a large spectrum of missense mutations (∼59%), frameshifts (∼15%) and splice site mutations (∼9%). FIX is a single-chain, vitamin-K dependent 57 kDa glycoprotein, synthesized by hepatocytes as a pre-pro-protein of 461 amino acids (aa) and it is structurally similar to both FX and FVII (Kah Yuen et al., 2017). As for Hemophilia A, the treatment option mainly relies on replacement therapy, based on purified, recombinant concentrates or Prothrombin Complex Concentrate (PCC). In contrast, the incidence of the other bleeding disorders is around 1:1,000,000 and 1:2,000,000 in the general population. Therefore, these are referred to as Rare Bleeding Disorders (RBDs) and are typically orphan diseases. Coagulation FVII deficiency (OMIM #227500) is recognized as the most common among the RBDs, with an estimated prevalence of symptomatic individuals of about 1:500,000 (Mariani and Bernardi, 2009; Napolitano et al, 2017).

FVII deficiency is a rare monogenic, recessive disease resulting from an heterogenous pattern of mutations in the F7 gene, located at 13q34. It is estimated that around 78% of the disease-associated mutations are missense mutations associated with reduced FVII plasma activity in affected individuals. FVII is a vitamin K-dependent serine protease produced by the liver parenchyma as a 50 kDa multidomain glycoprotein. Native FVII consists of 406 aa, composing a N-terminal γ-carboxyglutamic acid domain followed by two epidermal growth factor-like domains (EGF1 and EGF2), and a C-terminal protease domain (Monroe and Key, 2007). It is a unique factor that includes a fraction of around 3% circulating in an active form, even in the absence of coagulation activation. As previously mentioned, FVII plays a fundamental role in coagulation, initiating coagulation upon interaction with TF. FVII deficiency is characterized by a wide, heterogeneous pattern of clinical phenotypes, ranging from asymptomatic to severe and lethal hemorrhagic forms, even in the presence of the same levels of FVII coagulant activity. The most frequent clinical manifestations are mild, whereas 15% of the affected individuals exhibit severe, life-threatening bleeding episodes. The recurrent bleeding symptoms are similar to those presented in platelet disorders, with frequent epistaxis, menorrhagia, post-operative bleeding, and bruising. The severe disease phenotype is clinically defined as FVII activity below 10%, correlating with higher risk of recurrent and intense bleeding, while the moderate form, 10%–20% and the mild deficiency from 20% to 50% activity. In severe cases, Central Nervous System (CNS) and Gastrointestinal (GI) hemorrhages are relatively common, as well as joint bleeding and hemarthrosis-related arthropathy (Mariani et al., 2005). As for the other bleeding disorders, the current treatment is based on replacement therapy with highly expensive and frequent injections of exogenous FVII concentrates, either purified from plasma (pdFVII) or produced by recombinant technology (rFVII). Unfortunately, rFVII requires frequent bolus injections and careful monitoring due to its short half-life (<3 h).

As a consequence, two recent pioneering studies had the ambition to investigate HOs generated from Hemophilia patients where the disease-causing mutation was corrected using CRISPR/Cas9 technology. Their aim was not only to model the disease but also treat the disease-phenotype by long-term engraftment. Luce et al. (2021) reported the generation of HOs from a severe Hemophilia B (HB) patient, carrying the g.31280 G>A (c.1297G>A, p.E433K) mutation. In particular, this alteration leads to severe FIX inactivation, which is clinically characterized by FIXc <1% and FIXAg 67% levels respectively (Luce et al., 2021). Since conventional 2D differentiation displayed incomplete in vitro hepatic maturation, the patho-molecular mechanisms underlying the g.31280 G>A mutation have been investigated using a novel HO model. Consequently, HOs have proven a unique platform to investigate and uncover the molecular and intracellular mechanisms involved in disease pathogenesis, since they harbor the exact constellation of genetic variants present in the parental patient-derived iPSCs (Lancaster and Huch, 2019). Several studies have demonstrated the versatility of CRISPR/Cas9 in precisely editing exonic non-synonymous single nucleotide variations (nsSNVs) as well as large insertions, and this technology is particularly useful in the treatment of coagulation factor deficiencies (Lyu et al., 2018; Roper and Yilmaz, 2019; Son et al., 2022). Accordingly, syngeneic disease and non-disease models can be generated by either correcting coagulation deficiency-affected iPSCs or re-creating relevant point mutations, consequently overcoming the obstacle of sample accessibility and experimental variability (Hockemeyer and Jaenisch, 2016; McTague et al., 2021). In addition, the iPSC-based, CRISPR-edited system may allow the modification of a single allele, thereby recapitulating the genetic background of a heterozygote (Rabai et al., 2019; Prat et al., 2020). This strategy has been adopted by Luce et al. (2021) where they have generated syngeneic disease and disease-corrected models to investigate HB and to pave the way for the next experimental phases, consisting of gene editing and cell-therapy for HB. In additional studies performed by Lyu et al. (2018), it has been recently demonstrated that it is possible to stably restore FIX secretion upon integration of a therapeutic copy of F9, via the Adeno-Associated Virus Integration Site 1 (AAVS1) locus, using CRISPR-mediated Homology Directed Repair (HDR). Based on these findings, Luce et al. (2021) then moved far beyond the initial modeling purposes, by introducing a therapeutic F9 minigene into HB iPSCs carrying the Padua mutation (g.31134 G>T, p.R384L) to enhance FIX activity, prior to HO differentiation. When using standard 2D recombinant technology and culture system, Luce et al. (2021) reported dramatic and substantial differences in size, half-life and activity of the recombinant FIX protein. This was caused by complex, in vitro post-translational modifications, which were responsible for an abnormal lack of γ-carboxylation of the GLA-rich domain. However, in HOs where the mutation was corrected, the recombinant FIX protein was the same size, contained the expected post-translational modifications and global clotting activity of the plasmatic protein (Luce et al., 2021). For this reason, they transplanted the corrected HOs into newborn HB, F9 KO mice and monitored them for 2 weeks. They demonstrated that the F9 minigene knock-in could rescue FIX physiological levels and activity, thus reversing the disease phenotype in F9-KO mice. Interestingly, these preliminary findings overcame one of the relevant challenges presented in the previous studies of Ramaswamy et al. (2018), describing drastically limited therapeutic effects and compromised engraftment when using conventional 2D iPSC-derived hepatocyte-like Cells (HLCs). Luce et al. (2021) also highlight that single-cell techniques led to dramatically low performances in terms of hepatic maturation, proliferation, transplantation, and therapeutic relevance. As a consequence, they addressed the need of novel 3D approaches, to circumvent these limitations and give a long-term, therapeutic solution to HB. In this regard, the study is strongly limited by a short-termed experimental layout, and definitely requires long-term analysis in alternative animal models, such as immunocompromised, F9-KO mice.

Another pioneering study has been carried out by Son et al. (2022), achieving long-term therapeutic correction of Hemophilia A. As a proof-of-principle they have demonstrated that CRISPR/Cas9 technology could correct Inv22 in patient-derived iPSCs. Here, HOs from genome-edited iPSCs were transplanted into hemophilic mice, where they detected secretion of significant amounts of functional FVIII, followed by a dramatic reversion of the bleeding phenotype. Based on these findings, they hypothesized that the liver-like, multi-cellular repertoire of HOs could provide an optimal microenvironment for resident endothelial cells to boost and sustain corrected-FVIII expression over time (Son et al., 2022). In addition, they anticipated how multicellular 3D liver organoids could enable a more effective correction of HA when compared to traditional 2D techniques, by reducing the number of cells to engraft and enhancing the expression of functional FVIII (Son et al., 2022). They calculated that engraftment of ∼2800 HOs, containing 4*10^{^6} endothelial cells was sufficient to achieve a therapeutic outcome (Son et al., 2022). However, the study is limited to only 2 weeks evaluation, and by the use of immunocompetent HA mice treated with cyclosporine A.

As presented, recent studies have addressed the possibility to deliver genome-edited, patient-derived HOs through feasible, autologous transplantation and investigate these cells for their ability to revert the disease phenotype in a spectrum of bleeding disorders and liver diseases (Hockemeyer and Jaenisch, 2016; Son et al., 2022). Especially recent studies by Lyu et al. (2018), Ramaswamy et al. (2018), Luce et al. (2021) and Son et al. (2022) have laid the foundation for novel cell therapy to be potentially realized clinically for the treatment of coagulation bleeding disorders, in particular Hemophilia, in addition to liver disease (Figure 3). They have demonstrated how genome-edited, highly functional, iPSC-derived HOs can completely revert the bleeding phenotype of Hemophilic mice. However, these preliminary, pre-clinical achievements are limited to short-timed investigations, due to the lack of immunocompromised, KO mice. It is therefore an urgent need of studies exploring the therapeutic potential of HOs, and the survival and tolerogenic competency of the preclinical models in a long-term perspective. Several studies and clinical trials have described an unexpected role of hepatocytes and LSECs in modulating the immune response, by suppressing CD8⁺ and CD4⁺ T cell activation, and regulatory T cells (Tregs) (Arruda and Samelson-Jones, 2016; Crispe, 2016; Pratt et al., 2020). This may be one of the reasons behind the unprecedented therapeutic relevance of HOs, which is corroborated by the in vivo-like nature and hepatocellular heterogeneity of these cells. However, concerns regarding the safety of genome editing, particularly upon delivery in vivo or transplantation must be taken seriously. Thus, this is currently under major consideration, and several studies have recently demonstrated the effectiveness and, particularly, safety of genome editing in a broad spectrum of hematological and coagulation diseases (Hoban et al., 2016; Ou et al., 2016; Ohmori et al., 2017; Lyu et al., 2018; Ramaswamy et al., 2018; Cosenza et al., 2021; Son et al., 2022). Importantly, clinical trials implementing CRISPR-edited HSPCs for the treatment of severe Sickle Cell Disease (SCD) and β-thalassemia are currently on-going, with no reported off-target effects (Frangoul et al., 2021). In the case of β-thalassemia, the U.S. Food and Drug Administration approved Zynteglo^®, a one-time, single-dose gene therapy product to treat adult and pediatric β-thalassemia patients who require regular transfusions, in August 2022. With a prodigious 89% success rate, consisting of transfusion independence for at least 12 months, no signs of malignancies have been reported (FDA, 2022). The major concern of these treatments is that unintended, non-specific modifications may arise at locations other than the targeted site, due to sequence similarity (Zhang et al., 2015). However, several generations of Cas9 nucleases and accumulating protocols have maximized the HDR competency, while drastically abolishing residual side effects. In addition, extra precautionary measures are normally done where genome-edited iPSCs are characterized ex-vivo and examined for absence of off-target edits, homozygous or heterozygous correction occurrences, genome integrity and pluripotency, prior to differentiation toward HOs and transplantation (Jacków et al., 2019; Geng et al., 2020). This allows for the generation of safe, stepwise protocols to ensure and validate the long-term safety of the patients. When compared to the several limitations with standard 2D approaches, HOs have shown an exceptional ability to engraft into the kidney capsule of mice, either with or without ECM, and be maintained over time (Roper et al., 2017; Harrison et al., 2020; Son et al., 2022). The kidney is commonly considered the preferential administration route for HOs, since it is easily accessible and highly vascularized, and the retroperitoneal location allows advanced imaging and biopsy (Shultz et al., 2014). This method relies on the groundbreaking studies of Takebe et al. (2013), who was the first to establish iPSC-derived vascularized embryonic Liver Bud (iPSC-LB) organoids composed of Umbilical Cord Vein Cells (HUVECs), Mesenchymal Stem Cells (MSCs) and iPSC-derived hepatocytes. Following transplantation in immunocompromised mice, the vasculature of iPSC-LBs connected to the host vessels, thus allowing an efficient engraftment and the development of proliferating hepatoblasts into mature, highly functional hepatic tissue (Takebe et al., 2013). Consequently, this has demonstrated that it is possible to rescue drug-induced liver failure without the recipient liver replacement (Espejel et al., 2010; Takebe et al., 2013). However, obtaining a clinically relevant HO engraftment in humans is challenged by several considerations, particularly in terms of cell number and survival. Indeed, the minimal scale required for a potential clinical translation is generally identified as 10^{^9} hepatocytes per single adult patient (Horslen and Fox, 2004; Olsen et al., 2021). Most of the current approaches are still set up for the pre-clinical scale, and not yet optimized for the industrial scale and human application. In this perspective, the recent advances in increasingly optimized and standardized differentiation protocols, as presented by Harrison et al. (2020), and in the use of bioreactors may exponentially boost the establishment of an effective and safe bioprocessing, as well as future clinical trials. As presented by Luce et al. (2021), a possible alternative to refine the amount of cell needed for therapeutic purposes may be the use of strategies to enhance protein expression. It is critical to highlight that coagulation factor deficiencies can be corrected or improved from severe to mild clinical presentations even by minor increases in the protein levels and activity. However, there are many obstacles to overcome in order to clinically translate the potential of these novel iPSC- and HO-based therapies. Olsen et al. (2021) have observed that the in vitro differentiation of iPSCs to organoids correlate with low degrees of inflammation, which often results into batch variability and HSC-like cell stimulation. For this reason, a further optimization of the current culture conditions is needed. Furthermore, genomic stability, fluctuation in the reprogramming quality and functional pluripotency among iPSC lines, and residual iPSCs in the HOs to be transplanted should be taken into consideration (Son et al., 2022). To address this, increasing stepwise validation methods have been developed in order to examine the quality, the content, and the safety of HOs. While the genome integrity of iPSCs and, particularly, CRISPR-edited iPSCs is validated by whole genome DNA sequencing, HOs are carefully interrogated via proteomic analysis and single cell sequencing for the absence of any residual iPSC, to completely eradicate the risk of teratomas. Also, such analyses are needed to confirm the identity and quality of the cell subtypes present in the HOs, to ensure the safest and most efficient therapeutic benefit. Another significant limitation in cell therapy is the graft rejection, caused by the immune host reaction as a consequence of a non-compatible or partially mismatched Human Leukocyte Antigen (HLA), or Major Histocompatibility Complex (MHC) (Madrid et al., 2021). Consequently, patients’ iPSC-derived HOs are autologous and circumvent this obstacle, thus allowing for long-term engraftment without an adverse reaction or the need of immunosuppressive regimens. To investigate this further, several clinical trials for autologous iPSC-based transplantations have recently validated the absence of side-effects after 5 years of follow-up (Palta et al., 2014; Takagi et al., 2019; Schweitzer et al., 2020; Madrid et al., 2021). Numerous studies in the last decade have reported with success the complete absence of immune rejection after autologous iPSC-derived cell therapy in animal models (Araki et al., 2013; Guha et al., 2013; Osborn et al., 2020; Schweitzer et al., 2020). To overcome these challenges, further investigations are necessary and the use of larger animal models, such as pigs, are required in order to define the right dose per patient, the survival of engrafted HOs, and the long-term therapeutic outcome, and the tolerance of the transplant recipient, prior to clinical trial assessments expected in the coming years.