Therapeutic potential of hematopoietic cells derived from pluripotent stem cells

Abstract

Several challenges must be overcome before hematopoietic cells derived from pluripotent stem cells (PSCs) can be tested in the clinics. Pre-existing genetic mutations in somatic cells are a major concern for the production of iPSCs (induced pluripotent stem cells). This raises the question of what is the best somatic cell source to reprogram into iPSCs. Adult stem cells such as germ cell precursors and hematopoietic stem cells (HSCs) which are believed to be protected from somatic mutation accumulation are good candidates. Three gene editing methods have now been developed for human cells. Careful comparison of these methods is needed to determine the most appropriate for clinical applications. Differentiation of PSCs generally recapitulates early development. Therefore, cells produced from PSCs have an embryonic phenotype. Because transplantable HSCs and red blood cells expressing adult hemoglobin arise late in development, long after gastrulation, they have been difficult to produce from PSCs. The most difficult challenge is perhaps the development of methods to produce cells with an adult phenotype. Interestingly, recent reports suggest that primitive hematopoietic cells might make important contributions to adult hematopoiesis. Production of primitive hematopoietic cells might therefore have clinical applications.

Keywords::

• erythropoiesis
• hematopoiesis
• pluripotent stem cells
• temporal regulation

Isolation and culture of human embryonic stem (ES) cells was first described by the Thomson Lab in 1998. In 2006, Takahashi and Yamanaka reported that it was possible to reprogram somatic cells into induced pluripotent stem cells (iPSCs) resembling ES cells by overexpression of four multipotency factors [1]. Although iPSCs and ES cells can be subtly different, they are functionally equivalent and can be generically referred to as pluripotent stem cells (PSCs). These two seminal reports, augmented by the work of many labs, have provided the means to produce unlimited amounts of PSCs that can theoretically differentiate into any cell types. PSCs are therefore a potential source of therapeutic cells for regenerative medicine. In the hematopoietic system, some of the major applications are the production of mature fully differentiated cells such as red blood cells and platelets for transfusion [2] and immune cells such as natural killer (NK) cells [3] and antigen-presenting cells for cancer immunotherapy. Transplantable hematopoietic stem cells (HSCs) that could be used for all the existing transplant indications including cancer therapy are also a major target. Genetic modification of patient-specific iPSCs followed by the production of transplantable HSCs could provide an autologous cure for all genetically inherited disorders that arise from cells derived from the hematopoietic system. It has been proposed that even disorders such as hemophilia that do not originate in the hematopoietic system could be treated by genetic modification of transplantable HSCs [4].

Several challenges must be overcome before these therapies can be tested in the clinics. There has been considerable progress in the production of transgene-free iPSCs but several studies have shown that iPSCs can contain significant amounts of genetic and epigenetic abnormalities [5]. Importantly, it was shown that many of the genetic differences pre-existed the reprogramming and had been acquired by the somatic cells either before harvest or in culture. This raises the question of what is the best source of cells for reprogramming. Recent sequencing studies demonstrated that in humans there are about 30 – 50 de novo mutations when a genome is transmitted from one generation to the next [6]. This very low number of genetic abnormalities can be considered as the gold standard for cell therapy applications. These studies also suggested that there are powerful mechanisms in both the male and female germ lines that prevent the accumulation of mutations. Germ cells precursors are therefore a very attractive source of cells for reprogramming, although their harvest is invasive. Adult stem cells such as HSCs or muscle stem cells might harbor mechanisms that prevent the accumulation of mutations and are therefore also potentially acceptable somatic cells for the production of clinical grade PSCs.

One major difficulty is that the amount of genetic abnormalities that is acceptable for clinical use is unclear. Precise characterization of the genetic abnormalities that are present in hematopoietic cells that are harvested in vivo for allogeneic transplantations provides a useful baseline to determine the suitability of reprogrammed cells for clinical use [7].

There has been major progress in the ability to site-specifically genetically modify human cells. No less than three methods are now available: zinc-finger nuclease, TALEN and the newly developed CRISPR system [8]. All three methods are based on the creation of double-strand breaks to stimulate homologous recombination and potentially suffer from the same drawbacks. Careful optimization of the delivery methods and comparison of the efficiency and of the off-target cleavage rate of these methods must be performed in order to select the most appropriate system for regenerative medicine.

The greatest remaining challenge is perhaps the development of methods to differentiate PSCs into therapeutically useful cell types. Developmental hematopoiesis and erythropoiesis have been extensively studied and are well understood in the mouse but less well so in humans [9]. The globin switches are particularly useful to assess the developmental maturation of the cells produced by differentiation of PSCs because they have been very well characterized and because globin chain expression is easy to measure.

The first morphologically recognizable hematopoietic cells produced in human development are primitive erythroid cells and macrophages that are produced from week 3 to week 7 or 8 of gestation in the yolk sac. These cells are quite different from their adult version and their precursors are believed to lack long-term self-renewal potential since these cells are produced only transiently during development. At this stage, red blood cells express mostly embryonic globins. Starting at weeks 6 – 7 of gestation, primitive hematopoietic cells are progressively replaced by cells derived from self-renewing HSCs that are produced from hemogenic endothelium that originate in the aorta-gonado-mesonephros (AGM) region of the embryo. In the mouse, production of self-renewing HSC is preceded by a wave of definitive erythroid myeloid progenitors (EMP) that do not self-renew but that produce a wider diversity of hematopoietic cells than primitive hematopoiesis [10]. Concomitant with the emergence of AGM-derived cells, hemoglobin expression switches from embryonic to fetal. Maturation of the hematopoietic lineage does not end with the production of self-renewing HSCs, since in most hematopoietic lineages fetal cells differ significantly from cells produced in adults [11]. For instance, in humans, the type of hemoglobin produced switches from fetal to adult after birth.

This complex developmental history is likely the cause of the difficulties in producing transplantable HSCs from PSCs. Studies of globin expression clearly demonstrate that PSC differentiation recapitulates early development and suggest that the most mature red blood cells that can be produced in PSC differentiation cultures are equivalent to cells that would be found at week 5 – 7 of gestation [12]. It is likely that all cells produced from pluripotent cells have an embryonic phenotype. Cell types that are produced early in gastrulation are easier to obtain by differentiation of PSCs than cells that are produced later in development. In the mouse, budding of self-renewing HSCs occurs at day 10.5, but final specification might occur only after migration to the fetal liver at day 12.5, long after the end of gastrulation [13]. In order to understand the studies in greater detail at the molecular and cellular studies of these late developmental events are needed to help improve PSC differentiation protocol since it might be necessary to reproduce in vitro complex sequences of exposure to different niches and soluble factors to coax PSCs to differentiate into developmentally mature cell types.

A recent study in which fetal and adult erythroid progenitors were epigenetically profiled suggested that combinatorial assembly of developmental stage-specific enhancers was a critical determinant of developmental program and of temporal regulation of transcriptional networks [14]. Since expression of master transcription factors such as GATA1 and TAL1 was relatively unchanged throughout development, this study suggested that manipulation of the expression of the nuclear co-factors that control enhancer occupancy might be useful to control the developmental stage of maturation of cells produced in PSC culture. Transcription profiling of hematopoietic stem and progenitor cells at different stages of development and comparison with hematopoietic progenitor cells (HPC) produced from mouse ES cells revealed that genetically unmanipulated PSC-derived HPCs were most similar to yolk sac-derived cells [13]. Interestingly, HPCs produced from ES cells transduced with HoxB4 that have some self-renewal potential but abnormal differentiation potential were more similar to definitive than to primitive HSCs, suggesting that overexpression of HoxB4 incompletely re-specifies primitive cells to a definitive fate. This encouraging result suggests that it might be possible to completely re-specify PSC-derived cells to a definitive fate if the appropriate combination of factors is identified. Profiling studies on purified human HSCs and HPCs at different developmental stages are needed since the human and mouse hematopoietic systems differ by significant details. Mouse red cells for instance undergo only one major globin switch while human cells undergo two.

The optimal maturation stage of PSC-derived cells for clinical applications is unclear. As discussed in [2], in the case of red blood cells, adult hemoglobin expression would be preferable for most transfusion although HbF expression might be better for newborn transfusions and acceptable for other indications. Platelets derived from PSCs are embryo-like and larger than adult platelets. Additional functional characterization is needed to evaluate their clinical potential.

The contribution of yolk sac-derived primitive hematopoiesis to steady-state adult hematopoiesis remains unclear, particularly in humans. Interestingly, fate mapping experiments in the mouse revealed that postnatal hematopoietic progenitors do not significantly contribute to microglia homeostasis in the adult brain and that adult microglia derives from primitive myeloid progenitors that arise before embryonic day 8 suggesting that primitive cells derived from PSCs might have clinical applications in adults [15]. England et al. [16] reported that erythroblasts derived from mouse PSCs could divide in culture for much longer period of times than adult erythroblasts. The Paulson Lab has recently reported that stress erythropoiesis, a differentiation process that occurs in response to anemia and that rapidly produces red blood cells with fetal-like characteristics, is supported in the mouse by specialized erythroid-restricted progenitors that are transplantable and that can self-renew [17]. Although the developmental origin of these ‘erythroid-restricted' stem cells is unknown, it is tempting to speculate that they might arise from progenitors that emerge prior to HSCs during embryogenesis.

In conclusions, there has been considerable progress in the production, genetic modification and differentiation of PSCs. Additional basic science studies particularly with human cells, as well as additional translational studies must be completed before hematopoietic derivatives of PSCs can reach the clinic.

Declaration of interest

The author states no conflict of interest and has received no payment in preparation of this manuscript.