Directed Differentiation of Human Induced Pluripotent Stem Cells: A Retina Perspective

Abstract

The successful culture and characterization of human embryonic stem cells (hESCs) in 1998 [1] made unlimited human cell manufacture a plausible goal, ushering in a new era of science and technology, and promising to change the face of medicine. More than a decade of research later, significant advancements have been made in our understanding of the factors that govern hESC pluripotency and lineage-specific differentiation. Additional investigations into the safety and efficacy of transplanted hESC progeny in animal models of neural injury and retinal degenerative disease have brought the field to the brink of clinical trials. However, no sooner did the scientific, medical and lay communities become familiar with hESCs than a new source of pluripotent stem cells (PSCs) appeared on the scene. In 2007, the Yamanaka and Thomson laboratories independently published methods whereby human somatic cells could be reprogrammed to a pluripotent state by misexpressing a surprisingly small number of genes [2,3]. The resulting human induced PSCs (hiPSCs) appeared to have all the desirable attributes of hESCs without the ethical issues surrounding the use of human embryos. Furthermore, hiPSCs could be derived from individual patients, making it possible to develop customized stem cell therapies and generate disease-specific stem cell lines. Now that a number of laboratories have had the opportunity to evaluate both sources of PSCs, it is reasonable to examine the early findings, and ask how hiPSCs and hESCs have compared with regard to their capacity for directed cell differentiation. While several cell types could be used as readouts for such a comparison, the retinal lineage possesses a number of features that qualifies it as a robust and reliable barometer to assess the efficiency of targeted hESC and hiPSC differentiation. The retina has long served as a model of neural development owing to its accessibility, relatively simple architecture and conserved developmental program, as well as the availability of specific markers for some of its cellular constituents, such as photoreceptors and retinal pigment epithelium (RPE). Owing in large part to these advantages, a tremendous amount of information is available regarding the mechanisms, sequence of events and time course of vertebrate retinogenesis. Such knowledge is useful when evaluating the ability of PSC lines to differentiate along the retinal lineage, and, more importantly, for intervening when the process veers from the desired course. Numerous groups, including our own, have demonstrated that differentiating hESCs mimic the stepwise development of retinal cells in vivo [4–6]. Furthermore, hESCs appear to respond to secreted morphogens in a manner predicted by studies of vertebrate neural induction and retinogenesis. In particular, blockade of bone morphogenetic protein and canonical Wnt signaling is known to be important for neural and retinal patterning, and many retinal differentiation protocols call for antagonists of one or both of these pathways to be included in the culture medium [4,6]. Differentiation protocols that do not add bone morphogenetic protein or Wnt inhibitors likely rely upon their endogenous expression and secretion in culture, which could, at least partially, account for the observed ‘default’ production of early anterior neural and retinal cells [5]. In fact, numerous hESC lines, including the widely used H1 (WA01) and H9 (WA09) WiCell lines, generate anterior neuroectodermal derivatives with little, if any, manipulation of culture conditions. For example, overgrown hESC cultures will undergo spontaneous differentiation and generate patches of pigmented RPE, often in large abundance [7].