Abstract
Fanconi anemia (FA) is a rare inherited disease that is associated with bone marrow failure and a predisposition to cancer. Previous clinical trials emphasized the difficulties that accompany the use of gene therapy to treat bone marrow failure in patients with FA. Nevertheless, the discovery of new drugs that can efficiently mobilize hematopoietic stem cells (HSCs) and the development of optimized procedures for transducing HSCs, using safe, integrative vectors, markedly improved the efficiency by which the phenotype of hematopoietic repopulating cells from patients with FA can be corrected. In addition, these achievements allowed the demonstration of the in vivo proliferation advantage of gene-corrected FA repopulating cells in immunodeficient mice. Significantly, new gene therapy trials are currently ongoing to investigate the progressive restoration of hematopoiesis in patients with FA by gene-corrected autologous HSCs. Further experimental studies are focused on the ex vivo transduction of unpurified FA HSCs, using new pseudotyped vectors that have HSC tropism. Because of the resistance of some of these vectors to serum complement, new strategies for in vivo gene therapy for FA HSCs are in development. Finally, because of the rapid advancements in gene-editing techniques, correction of CD34+ cells isolated from patients with FA is now feasible, using gene-targeting strategies. Taken together, these advances indicate that gene therapy can soon be used as an efficient and safe alternative for the hematopoietic treatment of patients with FA.
Clinical Features of Fanconi Anemia
F
Molecular Biology of Fa
FA is a complex disorder characterized by marked chromosomal instability that arises due to defects in the repair of ICLs. Although exogenous agents that could generate ICLs (i.e., various chemotherapeutic agents such as cis-platinum) are known, endogenous sources that promote the formation of ICLs have remained elusive. Studies have confirmed that aldehydes generated from the metabolism of various molecules, including alcohols or fats, constitute important endogenous agents that can generate ICLs. 11 ICLs generated by either external agents or endogenous sources can be extremely damaging for FA cells, including hematopoietic stem cells (HSCs). 12 –15
FA is caused by mutations in any of the 22 genes that are known as FANC genes (FANCA, FANCB, FANCC, FANCD1/BRCA2, FANCD2, FANCE, FANCF, FANCG/XRCC9, FANCI, FANCJ/BRIP1, FANCL, FANCM, FANCN/PALB2, FANCO/RAD51C, FANCP/SLX4, FANCQ/ERCC4, FANCR/RAD51, FANCS/BRCA1, FANCT/UBE2T, FANCU/XRCC2, FANCV/REV7, and FANCW/RFWD3). 16 –21 In most cases the disease is autosomal recessive, with the exception of FANCB, which is X-linked, and mutations in the oligomerization domain of RAD51/FANCR, which are dominant negative. Mutations in FANCA occur in 60% of patients with FA, followed in frequency by mutations in FANCG and FANCC. 22,23 These FA proteins cooperate in the so-called FA/BRCA pathway, which functions mainly to sense lesions and coordinate ICL repair. 18,24
At a molecular level, FANCM is the first FA protein that, together with the FA-associated protein (FAAP24) and the DNA-binding cofactors MHF1 and MHF2, recognizes ICLs. 25,26 This recognition facilitates recruitment of other members of the FA core complex, including FANCA, FANCB, FANCC, FANCE, FANCG, FANCL, FANCT, FAAP100, MHF1, MHF2, FAAP20, and FAAP24. This complex can monoubiquitinate two different proteins, FANCD2 and FANCI, 27,28 through FANCL (ubiquitin ligase subunit) 29 and FANCT (UBE2T; E2 ubiquitin-conjugating enzyme). 17,19,20 After monoubiquitination, the FANCD2-FANCI complex moves to chromatin where recruitment of SLX4—that acts as a scaffold activating the MUS81-EME1, SLX1 and XPF(FANCQ)-ERCC1 heterodimer—facilitates the unhooking of the ICL lesion. 30,31 Then, translation synthesis polymerases bypass the lesion through insertion and extension of the opposite strand facilitated by REV1 and REV3-REV7 21 to generate a DNA duplex that can be used as a template for the homologous recombination (HR) of the double-strand break (DSB) in cooperation with the other FA proteins and canonical HR proteins. 16 When the ICL is resolved, the USP1/UAF1 complex deubiquitinates FANCD2-FANCI, allowing its reactivation when a new ICL is generated. 32,33
Somatic Mosaicism: a Natural in Vivo Gene Therapy in Patients With Fa
Spontaneous reversions of FA gene mutations have been identified in about 20% of patients with FA. Various mechanisms account for these genetic reversions including microdeletions, microinsertions, missense mutations, intragenic crossover, back mutations, and gene conversions. 1,34 –38 Although in most instances reversions occur in only one or a few specific cell types—mainly peripheral blood (PB) T cells—in a small proportion of patients with FA the same reversion was observed in various hematopoietic lineages, indicating that they occurred in pluripotent HSCs. 1,34 –38 The observation of progressive improvements in PB cell counts in a number of mosaic patients with FA evidenced the proliferative advantage of these corrected cells, 36 –38 suggesting that ex vivo gene therapy involving autologous FA HSCs could constitute a potential therapeutic approach for rescuing BMF in patients with FA.
Hematopoietic Transplantation in Patients With Fa
Although androgen therapies can promote transient increases in the number of PB cells, allogeneic hematopoietic stem cell transplantation (HSCT) constitutes the preferred therapy for BMF in patients with FA. 39 Because of the hypersensitivity of FA cells to DNA-damaging agents, 40 reduced conditioning regimens are currently used for HSCT in these patients. Although outcomes for HSCT from HLA-identical siblings were generally good, 41 –43 those for transplants using cells from alternative donors were initially much poorer. The inclusion of fludarabine in conditioning regimens 44,45 and T-cell depletion strategies in donor grafts resulted in marked improvements in the outcome of these HSCT modalities. 43,46 –48 Despite these advances, patients with FA who underwent HSCT showed an increased incidence of solid tumors—principally SCCs—likely due to the conditioning regimen and the occurrence of graft-versus-host disease (GVHD). 6,49,50
On the basis of the efficacy of hematopoietic gene therapy achieved for various monogenic diseases, 51 –55 and taking into account advances in preclinical and clinical gene therapy studies for FA, 56,57 this therapeutic approach is now considered to be an important alternative to HSCT for the restoration of hematopoietic function in patients with FA.
Gammaretroviral and Lentiviral Gene Therapy Studies in Mouse Models of Fa
Soon after the discovery of the first FA gene, FANCC, by Manuel Buchwald's group in 1992, 58 in vitro studies of gene therapy with gammaretroviral vectors (RVs) and adeno-associated vectors (AAVs) were carried out both in lymphoblast cells lines (LCLs) and hematopoietic progenitors from patients with FA-C. 59,60 Once the murine Fancc gene was cloned by the same team 61 and the first mouse models of FA-C were generated, 62,63 gene therapy studies in FA mice were initiated. Evidence of therapeutic efficacy was demonstrated using RVs, 64 –68 lentiviral vectors (LVs), 69 –74 and foamy viral vectors (FVs; Fig. 1). 75 Importantly, each of these vectors was tested in mouse models of FA, including Fanca –/–, 66,69,70,72 –74 Fancc –/–, 64,65,67 –69,76 and Fancd1/Brca2 –/–. 71 Because HSC defects in Fancd1 –/– mice closely resembled those observed in HSCs from patients with FA, 77 this model could be used to demonstrate that gene therapy confers a proliferative advantage of HSCs in vivo, 71 thus mimicking the behavior of reverted HSCs in mosaic patients with FA.

Advances in Fanconi anemia (FA) gene therapy since the mid-1990s. Shown are the main achievements obtained and future perspectives in FA gene therapy. BMT, bone marrow transplantation; CFCs, colony-forming cells; GT, gene therapy; hFA, human Fanconi anemia; HSCs, hematopoietic stem cells; HSCT, hematopoietic stem cell transplantation HSPCs, hematopoietic stem and progenitor cells; LV, lentiviral vector; mHSCs, murine hematopoietic stem cells; RV, retroviral vector. Color images available online at
Difficulties in the collection of HSCs from patients with FA were also reproduced in FA mouse models.
78
Thus, to improve HSC collection efficiency for gene therapy purposes, various drugs were combined with granulocyte colony-stimulating factor (G-CSF) to enhance its HSC mobilization potential. Significantly, the chemokine receptor antagonist AMD3100 was shown to synergize with G-CSF, resulting in significant mobilization of HSCs in two FA mouse models.
79
Similarly, an interleukin-8-related chemoattractant protein also enhanced the efficiency of G-CSF-mediated mobilization of HSCs from Fancg
–/– mice.
80
These studies, together with data obtained from patients who had a poor HSC reserve,
81
and more recently in an FA clinical trial (
To enhance the homing of corrected HSCs in transplanted patients, transduction protocols for FA HSCs have been significantly modified. In this respect, studies carried out in various FA mouse models showed improvements in the engraftment of HSCs that had been transduced for short periods of time. 72,75,82 In addition to improved engraftment, safety studies carried out in Fanca –/– mice transplanted with syngeneic HSCs transduced with the PGK-FANCA.Wpre* LV 83 confirmed the efficiency of this approach and showed the safety of LV-mediated gene therapy in an FA-A mouse model. 73
Myeloablative or submyeloablative conditioning constitutes a standard method to facilitate HSC engraftment in FA mouse models. Although conditioning may not be essential for the gene therapy of patients with FA, the mild hematopoietic phenotype observed in most FA models requires the depletion of endogenous hematopoiesis to facilitate engraftment of transplanted HSCs. In most instances, ionizing radiation was used for conditioning, although cyclophosphamide has also been employed to develop gene therapy approaches that have greater clinical applicability. 65,84 Because of the hypersensitivity of FA cells to cytokines such as tumor necrosis factor (TNF)-α 68 or interferon-γ, 85,86 these drugs have also been considered for use as conditioning agents. 85,86 More recently, nongenotoxic drugs, such as monoclonal antibodies targeting HSCs, have been used to condition Fanca–/– mice. 87 This latter approach would constitute an ideal conditioning method for FA gene therapy, should conditioning be necessary to treat these patients.
An alternative approach for facilitating engraftment of corrected FA HSCs involves accessory cells. In this respect, previous studies showed that intrabone transplantation of mesenchymal stem cells (MSCs) facilitated homing of transplanted HSCs from wild-type animals. 88 A more recent study demonstrated that coinfusion of MSCs with corrected Fanca –/– HSCs facilitates the engraftment of these cells in transplanted animals, 74 suggesting that MSCs may constitute relevant cell populations that could be considered for inclusion in clinical FA gene therapy strategies.
Experience From Previous Fa Gene Therapy Trials Involving Gammaretroviral Vectors
In contrast to the successes achieved in gene therapy for primary immunodeficiencies and hemoglobinopathies, 52 –55 to date clinical trials involving patients with FA have not shown engraftment of gene-corrected HSCs or clinical efficacy for reverting BMF. 89 –91 The first trial included three children and one adult with FA-C. 89 HSCs were obtained from BM or PB isolated from these patients after mobilization with G-CSF. Purified CD34+ cells were then transduced over 3 days with an RV carrying FANCC. Up to four infusions of corrected cells were given to these patients at intervals of 3 to 4 months. Although transient improvements in PB cell counts were noted, no sustained hematologic responses were observed, and no gene-corrected cells were observed several months postinfusion. Strikingly, in one patient who received radiation therapy for a concurrent gynecological malignancy, cells with the FANCC transgene could be detected.
A subsequent clinical trial was conducted in patients with FA-A who had not developed BMF. 91 CD34+ cells were prestimulated for 3 days, followed by two rounds of transduction with an RV encoding FANCA. No gene-marked cells could be detected in any of the enrolled patients beyond 3 months postinfusion, and only transient improvements in PB cell numbers were observed.
Possible reasons for the defective engraftment of corrected HSCs in these gene therapy trials include defects in the transduction of true HSCs after relatively long transductions with RVs, infusion of limited numbers of transplanted HSCs, or absence of patient conditioning before cell infusion.
Advances in Lentiviral Transduction of Fa Hematopoietic Stem Cells in Xenogeneic Transplant Models
On the basis of the conclusions obtained from previous FA gene therapy trials and studies in FA mouse models, gene therapy approaches for patients with FA have been progressively optimized. Given the efficacy and safety of LVs relative to first-generation RVs, two similar PGK-FANCA.Wpre* LVs were developed. 83,92 Both vectors were used to transduce FA HSCs in a short period of time under conditions that protect FA cells from oxidative damage, including the addition of N-acetylcysteine to the culture medium and maintenance of a low-oxygen atmosphere during cell culture. 83,92 –94 Because of the hypersensitivity of FA HSCs to TNF-α 95 transduction, the culture medium also included the TNF receptor-Fc fusion protein etanercept to prevent cytotoxic effects induced by this growth factor. 94 As deduced from data obtained in clonogenic assays, these conditions showed optimized survival of hematopoietic progenitors in patients with FA.
To study the functional properties of human HSCs, in vivo analyses of the ability of these cells to repopulate in immunodeficient mice are frequently performed. NSG (NOD/LtSz-scidIl2rg–/–) mice that lack T, B, and natural killer cells currently constitute one of the most efficient models to demonstrate human HSC functionality. 96 –98 In the case of FA, previous studies showed a lower hematopoietic stem and progenitor cell (HSPC) content compared with healthy donor HSCs, and also evidenced homing defects when these cells were transplanted into immunodeficient mice. 99 These observations, together with the limited number of HSCs present in the BM of patients with FA, 100 –102 have limited the study of the long-term repopulation capacity of these cells.
Although several studies showed the feasibility of correcting the phenotype of hematopoietic progenitors from patients with FA (i.e., reduced mitomycin C [MMC] hypersensitivity of FA colony-forming cells [CFCs]),
59,83,92,93,103
evidence of correction in FA repopulating cells has remained elusive. One study showed the presence of a low number of marked cells in a single transplanted mouse after the long-term culture of FA transduced cells,
104
an approach that is now not recommended for FA cells. Another study has demonstrated reproducible engraftment of corrected HSCs from patients with FA, using a clinically applicable transduction protocol.
94
In this study, CD34+ cells from patients with FA-A were mobilized with G-CSF and plerixafor, and then transduced for a short period of time with the therapeutic PGK-FANCA.Wpre* LV.
83
Notably, human myeloid and lymphoid cells were identified in transplanted mice, suggesting the engraftment of repopulating cells having multipotent differentiation capacity. Moreover, the observation of a marked increase in MMC resistance of engrafted progenitor cells demonstrated for the first time the phenotypic correction and in vivo proliferative advantage of corrected HSCs in patients with FA.
94
These observations thus suggested that a similar proliferation advantage of corrected FA HSCs should occur after infusion in patients with FA, ideally in the absence of conditioning. On the basis of improvements achieved in these experimental approaches, two gene therapy trials with FANCA LVs are ongoing with the aim of demonstrating phenotypic hematopoietic correction in patients with FA
57
(
Fa Gene Therapy With Nonviral Vectors and New Pseudotyped Lentiviral Vectors
On the basis of the safe integration profile of Sleeping Beauty transposon vectors, 105 this family of nonviral vectors was proposed for gene therapy approaches in various tissues and diseases that affect hematopoietic and nonhematopoietic tissues. 106 Improvements in the transposition of human HSCs 107 suggest that this simple and relatively inexpensive gene therapy approach will have a significant role in future therapies for hematopoietic diseases, including FA. 108
Because a current drawback in FA gene therapy derives from the difficulties involved in efficiently selecting CD34+ cells from patients with FA, several studies have pursued the targeting of HSCs using new pseudotyped vectors, such as the measles virus glycoprotein-pseudotyped LV (hemagglutinin and fusion protein LVs [H/F-LVs]) that efficiently transduce nonpurified HSPCs from patients with FA. 109 Conclusions reached by the use of these vectors in experimental models of gene therapy strongly support the relevance of their production under GMP conditions for clinical applications. In addition to these new ex vivo strategies of gene therapy, the in vivo transduction of FA HSCs also constitutes an attractive approach to avoid the need for in vitro manipulation of these sensitive cells. Because H/F-LVs are resistant to serum complement, these pseudotyped LVs would also constitute a good alternative for future approaches in FA gene therapy. 110 With the same objective, studies have shown the possibility of in vivo transduction of HSCs with adenoviral and foamy viral vectors, 111,112 suggesting that these new in vivo approaches would also be relevant to the development of future FA gene therapies.
Next Generation of Fa Gene Therapies Based On Gene Editing
To increase the safety of gene therapy, significant advances have also been achieved in gene targeting, due to the design of specific nucleases that markedly increase the efficiency of homologous recombination (HR) in defined sequences of the genome. The use of these nucleases combined with the transfer of donor DNA templates facilitated the correction of specific mutations, as well as the insertion of wild-type genes in safe harbor loci or immediately downstream of endogenous regulatory sequences (knock-in strategies). 113
Although the role of certain FA proteins such as BRCA1 and BRCA2 in HR is clear, the relevance of other FA genes in DNA repair processes is less certain. 114 Thus, the question of whether or not gene editing in FA cells could be feasible has long remained unclear. However, initial studies carried out with fibroblasts from patients with FA-A 115 as well as from patients with FA-C and FA-G 116,117 showed the possibility of using gene editing for FA cells. Significantly, a more recent study also confirmed the correction of specific mutations in primary fibroblasts from a BRCA2-deficient (FA-D1) patient. 118 Once gene editing was shown to be feasible in FA cells, editing in LCLs and CD34+ cells from patients with FA-A was demonstrated. 119 Designed zinc finger nucleases (ZFNs) and a PGK-FANCA sequence flanked by the AAVS1 homology arms targeted the FANCA gene at the AAVS1 locus with efficiencies of up to 10%. 119 Although the efficiency of gene editing in hCD34+ cells from healthy donors has markedly increased with the use of improved editing tools, 113 whether gene-edited FA CD34+ cells will also have in vivo repopulating ability, as was already demonstrated in lentivirally transduced FA-A HSCs, is unknown.
Finally, as was the case with conventional gene therapy, safety concerns must be considered for the editing of FA HSCs, particularly regarding the potential off-target activity of nucleases used in these approaches. Although ZFNs used in AAVS1-targeting studies showed no off-target activity in any of the top five off-target sites found in in silico analyses, 119 additional studies will be required before the use of these approaches in clinical FA gene therapy.
Advantages and Current Limitations of Cell Reprogramming in Fa Gene Therapy
Soon after cell reprogramming was described, 120 various groups began to use these techniques in regenerative medicine. Given that one aspect that limited hematopoietic gene therapy for FA was the low number of HSCs present in the BM of patients with FA, the possibility of generating these cells from other, nonhematopoietic tissues was thus considered. Theoretically, two alternatives can be used for this purpose. Cell types such as skin fibroblasts or keratinocytes could be reprogrammed to generate induced pluripotent stem cells (iPSCs). These pluripotent cells could be then differentiated toward the hematopoietic lineage and the FA genetic defect corrected afterward. An alternative approach focuses on genetic correction of nonhematopoietic FA cells, followed by reprogramming and redifferentiation of the cells. Previous studies showed that stable FA iPSCs could not be generated from noncorrected FA cells, indicating the involvement of the FA pathway in cell reprogramming. 121 One of the first events that occur during the acquisition of pluripotency is the generation of DNA damage, 122 as shown in various studies that found low reprogramming efficiency of uncorrected FA cells because of the inherent DNA repair defects of these cells. Alternative strategies were then used to increase the efficiency of FA cell reprogramming. 123 –129 In contrast to these approaches, gene complementation overcame the technical difficulties in generating bona fide iPSCs from patients with FA 115,121,130 and FA mouse models. 131,132
Despite the great potential of cell reprogramming in regenerative medicine, applications of these technologies in HSCT are currently limited by difficulties in repopulating hematopoietic cells in transplanted animals with either human or mouse iPSC-derived hematopoietic cells. 121,131 Although some studies have shown the possibility of engrafting hematopoiesis in mice, using reprogrammed cells, 133 further improvements are required before considering the use of these techniques for transplantation of patients with monogenic diseases affecting the hematopoietic system, such as FA.
Concluding Remarks
There have been marked advances in the development of efficient and safe gene therapy approaches for the treatment of various monogenic diseases. Here we have summarized the significant achievements and perspectives related to gene therapy involving FA HSCs (Fig. 1). In 1999, 5 years after the first proof of concept for RV-mediated gene therapy using FA cells was described, the first gene therapy trials involving these vectors in BM and G-CSF-mobilized CD34+ cells were conducted. Since 2002, the efficacy of LVs for the treatment of HSCs, either from FA animal models or patients with FA, was demonstrated. Since 2015, improved LV transduction protocols have been developed, and evidence of in vivo proliferative advantages of gene-corrected FA HSCs was demonstrated in xenogeneic transplantation models. These observations suggested that gene-corrected FA HSCs may progressively replace hematopoiesis in patients with FA treated with gene therapy, even in the absence of conditioning. Thus, the first clinical trials with LVs were initiated with the hope that these low-toxicity therapies based on the reinfusion of autologous cells will soon be a true alternative to allogeneic transplantation. New ex vivo and in vivo gene therapy approaches with LVs pseudotyped to specifically target HSCs, or based on gene editing, have also been proposed, and may provide additional avenues for the development of gene therapies for FA and other monogenic diseases that affect the hematopoietic system.
After the submission of this review, a new paper describing a novel lineage depletion system for the gene therapy of Fanconi anemia patients was published. 134
Footnotes
Acknowledgments
The authors are indebted to the patients with FA, their families, and clinicians from the Fundación Anemia de Fanconi for invaluable support. The authors also acknowledge the commitment of the Fanconi Anemia Research Foundation (FARF), the International FA Gene Therapy Working Group, and the Fanconi Hope Foundation for support in the advancement of FA gene therapy. FA gene therapy studies at the CIEMAT are supported by grants from the 7th Framework Program European Commission (HEALTH-F5-2012-305421; EUROFANCOLEN), Ministerio de Sanidad, Servicios Sociales e Igualdad (EC11/060 and EC11/550), Ministerio de Economía, Comercio y Competitividad y Fondo Europeo de Desarrollo Regional (FEDER) (SAF2015-68073-R), and Fondo de Investigaciones Sanitarias, Instituto de Salud Carlos III (RD12/0019/0023). Current FA gene therapy studies at the CIEMAT are also supported by Rocket Pharmaceuticals, Ltd.
Author Disclosure
The Hematopoietic Innovative Therapies Division receives funding from Rocket Pharma and has licensed the PGK-FANCA-Wpre* LV to Rocket Pharmaceuticals. J.A.B. is a consultant for Rocket Pharmaceuticals. P.R. and S.N. have no competing financial interests.
