Genetic Correction of Hematopoiesis in Fanconi Anemia: The Case for a Non-HSC-Autonomous Defect

Hematopoietic stem cells (HSCs), by virtue of their replicative and developmental potency, as well as relative accessibility, are an intuitive high-value target for therapeutic modification by integrating retroviral vectors. Even a cursory literature review combining the search terms “hematopoietic stem cell,” “retrovirus,” and “efficiency” will yield hundreds of published articles describing experimental protocols suitable for stable gene delivery. Many of those describe advances that have been driven by gains in understanding of viral-vector biology and the resultant clever engineering of particle and payload (see review¹). Yet, as highlighted by the work of Jacome and colleagues recently published in Molecular Therapy, this has not translated into the projected therapeutic efficacy for some applications—here, Fanconi anemia (FA).² By allowing instructive observations in FA patients to guide their experimental strategy, the authors thoroughly revisited a key player: the target cell and its “natural habitat.”

FA, after all, is a multisystem disorder with a complex genetic basis and prominent hematopoietic manifestations.³ The HSC pool in the bone marrow appears depleted in patients at diagnosis, and mobilization of potential target cells for genetic correction to the peripheral blood has been problematic.⁴ Although the quantitative stem cell deficit and its consequences relate directly to patient morbidity, they cannot (yet) be therapeutically addressed, except through allogeneic stem cell transplantation. Jacome and colleagues therefore took an alternative approach, focusing on the underlying qualitative deficiencies and their impact on the success of ex vivo gene replacement approaches in FA. The FA HSC phenotype comprises reduced progenitor clonogenicity, exaggerated sensitivity to reactive oxygen species (ROS), and repopulation defects in murine models. This is generally consistent with a role for FA proteins in maintaining DNA integrity, mediating inflammatory signaling, and balancing proliferation and survival.

Intriguingly, many of these phenotypes can be corrected by retroviral expression of complementary DNA sequences.^4,5 Why, then, is success in clinical trials so elusive for FA?^4,6 Agreement exists that ex vivo culture is generally detrimental to the retention of stem cell properties, but in FA attrition seems further compounded by the apoptotic propensity of targeted cells. Maintaining complete functional competency (i.e., “stemness”) during in vitro transduction culture while avoiding the myeloproliferative transformation for FA HSCs is no mean feat.^7,8 Moreover, although the full scope of the hematopoietic phenotype in FA remains elusive, Jacome and colleagues present evidence that careful attention to these “idiosyncrasies” will pay off in an effort to improve transduction efficacy and, hopefully, therapeutic benefit.

Realizing the potential for loss of stem cell potency and viability, the authors undertook several key maneuvers to minimize ex vivo manipulation. They took full advantage of a chimeric gibbon ape leukemia virus–amphotropic envelope protein (GALV-TR) to pseudotype vectors, and substantially reduced ex vivo transduction culture duration to 16 hours. Using lentivectors derived from HIV-1, such relatively brief incubations have been shown to produce substantial levels of chimerism from gene-modified cells in syngeneic murine models.⁹ This approach is particularly attractive in FA, in which the extended culture of hematopoietic cells exposed to vector, but not successfully transduced, can induce myeloproliferative transformation, at least in murine models.⁷ Brief exposures would seem to put a premium on optimizing ex vivo transduction efficiency by enrichment of CD34-expressing progenitor populations containing the stem cell targets (implicitly increasing vector particle–to–target cell ratio). However, the authors demonstrate that there is in fact very poor correlation of CD34 cell content among primary FA patient bone marrow samples with the hematological status of the patients, or in vitro progenitor clonogenicity.

Although their hypothesis that the CD34 immunophenotype does not reliably identify clonogenic progenitors and stem cells in FA patients awaits broadened validation, it led the authors to minimize graft manipulation and retain CD34^— “accessory cell populations” during ex vivo culture. They demonstrate to striking effect that the transduction of erythrocyte-depleted but otherwise unprocessed whole bone marrow can preserve cell viability and transduction rates among progenitor populations.¹⁰ FA progenitor and stem cells display an exaggerated sensitivity to ROS following cycles of reoxygenation.⁸ Here, lowering oxygen concentration in culture to physiological levels helped the investigators to further preserve viability without apparent compromise in target transduction. Taken together, their strategy yielded impressive preclinical results in primary FA patient marrow cells, including freshly thawed, previously cryopreserved samples. Readers will note limitations of the work, including a predominant reliance on in vitro assays, the relatively small number of patient sample xenografts, the focus on a single complementation group, and the open question of the extent to which the transduction protocol can be modified for use of γ-retrovirus vectors or scaled for clinical applications.

Nonetheless, the investigators go on to demonstrate that bone marrow stromal (mesenchymal) cells are retained and functionally corrected during transduction culture, leading them to speculate on the facilitative effect that such cells might have on hematopoietic reconstitution. The study does not offer mechanistic insight into the role of accessory cells in such a process. However, the investigators' hypothesis is eminently testable by combining graft engineering and transplantation strategies in murine models, and arguably presents the key conceptual departure in their present work. The concept that the correction of nonhematopoietic elements might improve reconstitution with genetically corrected cells in FA gains further credence from a recent article by Li and colleagues.¹¹ Those investigators showed that in situ injection of wild-type, or vector-transduced, mesenchymal stem cells in a murine model of FA with homozygous transgenic disruption of Fancg improves engraftment, cellularity, and clonogenicity from subsequently grafted wild-type hematopoietic cells.

Is the hematopoietic defect in FA HSCs fully explained by dysregulated apoptotic signaling, oxygen radical sensitivity, or even DNA repair defects? Moreover, is the progressive exhaustion of hematopoiesis truly accounted for by an isolated functional HSC deficiency? Allogeneic stem cell transplantation can cure FA patients of the progressive marrow aplasia, thereby invoking a stem cell–specific deficiency phenotype. However, because bone marrow and even CD34-enriched grafts contain abundant accessory cells, this alone does not rule out a stromal defect. Indeed, a conclusive demonstration of the hypothetical selective advantage gained from functional (retroviral) correction that would follow from an isolated FA HSC defect is lacking in clinical trials and most murine models, in the absence of additional selective pressure. Can we dissect replicative and survival deficiencies from those potentially affecting progenitor and stem cell homing, migration, adhesion, and retention in the hematopoietic niche? Gene therapy studies in syngeneic murine models, and certainly the limited FA patient experience, might support a less restrictive interpretation involving not only HSCs but the hematopoietic microenvironment. Non-FA models indicate that hematopoiesis is not a cell-autonomous process and that the regulation of sustained multipotency and self-renewal requires the coordinate engagement of nonhematopoietic elements within the bone marrow microenvironment.¹² Thus, to attain a full understanding of the hematopoietic defect in FA, several important questions remain to be addressed (Figure 1):

Figure 1.

Potential mechanisms involved in the hematopoietic phenotype in Fanconi anemia. (1) Adhesion and anchorage of quiescent hematopoietic stem cells (HSCs).^12,13 (2) Trophic and structural support of HSCs by stromal and osteoblastic elements.^3,15 (3) Polarity, asymmetric cell division, and self-renewal capacity.^16,17 (4) Homing and migration of HSCs.^17,19 cdc42, Rho GTPase; G₀, cell cycle phase; IFN-γ, interferon-γ; MMP-7, matrix metalloproteinase 7; p38, p38 mitogen-activated protein kinase; ROS, reactive oxygen species; STAT5, signal transducer and activator of transcription 5; TNF-α, tumor necrosis factor-α.

1. Is the bone marrow microenvironment in FA fully competent for adhesion and anchorage of quiescent stem cells to support long-term hematopoiesis from genetically corrected HSCs?¹³ Others have previously proposed that the capacity for homing to and retention in the niche are prerequisites for sustained quiescence (G₀) and avoidance of the inflammatory, more oxygen-rich environment in which activated cells reside.¹² Incidentally, mice with a disrupted Fancc reading frame have been found to have fewer highly purified stem/progenitor cells in G₀ and a lower proportion of cells of the most immature immunophenotype.¹⁴

2. In a related matter, do stromal and osteoblastic elements present in the hematopoietic niche provide trophic and structural support conducive to sustaining hematopoietic cells in quiescence, or does the microenvironment in FA even actively contribute to senescence and stem cell attrition, for example, through the production of inflammatory mediators?^3,15

3. The generation of cell polarity affects the proliferative and regenerative capacity of HSCs through its role in asymmetric cell division. Can FA cells (HSCs) appropriately polarize to allow asymmetric division and thereby support self-renewal?^16,17

4. The notably poor mobilization of CD34⁺ cells from FA patients (and Fancc mice) with granulocyte colony-stimulating factor but relatively good response to AMD3100 (in mice) denotes disease-associated dysregulation.^4,18 Along with additional recent studies, this raises further questions regarding the capacity of FA HSCs to appropriately migrate, home, and lodge after ex vivo correction and intravenous injection.^17,19

Understanding and accommodating the FA HSC phenotype to improve (gene) therapeutic strategies will require a systematic analysis and perhaps even deliberate manipulation of the hematopoietic microenvironment.²⁰ This will take close collaboration among investigators based in developmental and, in particular, stem cell biology, as much as among those with expertise in vector engineering. The work by Jacome and colleagues nicely illustrates this shift in paradigm for targeting FA stem cells, but in a larger sense it also reflects the fact that cell and gene therapy do not exist in isolation. Success for FA gene therapy is intimately linked to key questions in stem cell biology, including polarity and HSC anchorage, as well as trophic and structural support in the microenvironment. As the field emerges from a thorough evaluation of retroviral genotoxicity, revisiting (noninduced) target cell biology will bring renewed energy and talent to a society, ASGCT, that aptly pursues a broadened focus.