ES cells, derived from the inner cell mass of blastocyst-stage embryos [57
], are able to maintain an undifferentiated state or can be directed to mature along lineages deriving from all three germ layers—ectoderm, endoderm and mesoderm. Photoreceptor features were found in subretinal human ES cell grafts, but not in locations elsewhere in the eye, indicating that the subretinal niche may be critical to support differentiation of ES cells towards a photoreceptor fate [60
]. Retinal fate has also been induced in mouse, monkey and human ES cells by using growth factors, retinal co-culture and genetic modification [61
]. When directed to become retinal precursors similar to the human foetal stage, human ES cells were found to integrate into an explant model of Leber congenital amaurosis [63
] and restore some function in vivo
]. These data strongly support the use of precursor cells for photoreceptor replacement. It may be that for effective integration, stem cells need to be differentiated some way along the photoreceptor lineage before transplantation; recent evidence has suggested that even mature neurons may integrate [65
]. Very recently, it was shown that three-dimensional culture of mouse ES cell aggregates led to autonomous optic cup formation in vitro
with features of retinal stratification [66
]. A similar system might be applicable to expand and differentiate a single ES cell into potentially thousands of photoreceptor precursors (or some other photoreceptor developmental stage) for optimal integration.
Despite the relative ocular immune privilege, ES cells are immunogenic as they originate from another human foetus, which also raises ethical questions about how to source these cells. The discovery that reprogramming DNA-binding proteins may induce stem cells from adult cells represents a milestone in the search for a renewable source of cells [67
]. These ‘induced pluripotent’ stem (iPS) cells, being autologous, may obviate the need for chronic immune suppression.
iPS clones could be derived from patients and be used for treatment—a process that may involve ex vivo
correction of the gene defect before reintroduction into the host. Recent progress has been made as iPS cells derived from amyotrophic lateral sclerosis patients have been differentiated into motor neurons—the cell type that requires replacement in this condition [70
]. In a similar vein, human iPS cells have been cloned from Parkinson's disease patients [71
] and iPS cell-derived dopamine neurons have improved function in a Parkinson's disease model [72
]. However, iPS clones vary in pluripotency and differentiate less efficiently than ES cells, which show robust neuronal differentiation [73
]. Interestingly, this variability is independent of the type of vector used in iPS cell production. Integrating vectors such as lentiviruses (whereby genes are inserted into the target cell genome), in addition to affecting pluripotency, confer a greater potential risk of teratogenicity than non-integrating vectors (whereby the gene is expressed while remaining separate from the host genome). By avoiding the use of genes and vectors associated with uncontrolled proliferation [74
], the risk of tumour formation is reduced. Ideally, iPS cells derived from every patient will need to be screened for potentially cancerous cells, as even a 0.01 per cent risk of malignancy induced by therapy may be unacceptable for patients and doctors. The situation is somewhat different with ES cells, which are not derived from individual patients—potentially, a few well-characterized, purified and approved cell lines may be used widely. Overall, the regulatory environment will be complex, given the range of pluripotency, differentiation capacity, teratogenicity and immunogenicity of different iPS cell clones and ES cell lines. It is likely that a range of pre-treatment protocols, potentially subject to differing regulatory requirements, will be developed in future, tailored to specific clinical situations.
iPS cells have differentiated in vitro
into retinal cell phenotypes, recapitulating events in normal development [76
]. Functional human photoreceptor-like cells have been observed following differentiation of human iPS cells [78
]. In the case of retinal degenerations, iPS cells from patients might be used for replacement, with correction of the RP gene defect and the cells then directed to assume a retinal precursor fate before transplantation (). For therapy, abundant autologous cells should be generated at a stage with maximal integrative capacity. For regulatory purposes, it would be optimal to use protocols that do not rely upon materials of bacterial or animal origin; low-molecular-weight compounds have been shown to induce retinal progenitors, RPE and photoreceptors from ES and iPS cells [80
Figure 4. 1. A suitable cell is obtained from the patient with retinal disease. An easily accessible source of cells is chosen so as to minimize surgical trauma. The example of a skin fibroblast is depicted here. 2. The capacity for self-renewal—absent (more ...)