Generation of Defined Astrocytic Phenotypesfrom Human Pluripotent Stem Cellsfor Transplantation after Spinal Cord Injury

Abstract

Spinal cord injury (SCI) is characterized by axonal damage, neural degeneration, formation of cystic cavities, and upregulation of a plethora of inhibitory as well as inflammatory molecules. To protect the surrounding tissue from further damage, fibroblasts and reactive astrocytes form an impenetrable barrier lining the lesion site. This environment impedes endogenous regeneration of axotomized neurons and glial cells. In addition to being exposed to extrinsic inhibitors of axonal regeneration after SCI, neurons in the mammalian central nervous system (CNS) intrinsically lack the capacity for spontaneous axonal regeneration. To restore neural tissue integrity and provide a favorable environment for injured axons to regenerate, an appropriate substrate is needed. An optimal substrate for neuroregeneration should have a neural identity and fulfill several functions including physical support and guidance, trophic and metabolic support, maintenance of tissue homeostasis, and modulation of neuronal outgrowth and network activity. Astrocytes represent the most suitable cell population to fulfill this task. Induced pluripotent stem cells (iPSCs), which show many homologies to embryonic stem cells (ESCs), represent an ethically acceptable means to obtain large amounts of astrocytes in vitro for autologous transplantation. In the present study, populations of immature astrocytes with caudal identity were generated from three human pluripotent stem cell (PSC) lines, whereby terminal in vitro differentiation was performed according to previously published studies using the inductors of astrocytic maturation CNTF, BMP2/4 and FGF1 to allow for comparative analysis of specific astrocytic subtypes and for selection of a pro-regenerative cell population. In accordance with previously published findings, morphological and functional differences among astrocytic populations were observed in the present study. Compared to astrocytes differentiated with FBS only, astrocytes differentiated with BMP2/4 exhibited a significantly increased cell size and a complex cytoarchitecture, decreased expression of the NSC marker Sox2, increased production of potentially growth-inhibitory extracellular matrix (ECM) components and of the growth-promoting neurotrophic factor BDNF, and limited ability to induce neurite outgrowth in co-cultures with primary dorsal root ganglion (DRG) neurons. Astrocytes differentiated with FGF1 were, in contrast, significantly smaller, mainly had a bipolar morphology and retained expression of the NSC marker Sox2, indicative of a rather immature phenotype. On the other side, differentiation with FBS alone or in combination with CNTF led to astrocytes that produced the pro-regenerative ECM component laminin, which was reflected in a strong growth-promoting effect on primary DRG neurons. Importantly, clear differences across the three used pluripotent stem cell (PSC) lines were observed; in particular, differences in the electrophysiological response pattern elicited by stimulation with adenosine triphosphate (ATP) indicated that astrocytic lines with a similar phenotypic profile have distinct characteristics. After transplantation into the intact and injured spinal cord of Fischer 344 rats,some of these phenotypical characteristics observed in vitro were maintained in vivo. This is indicative of a stable phenotype of pre-differentiated astrocytes, which allows the selection of an astrocytic subtype in vitro to maximize the pro-regenerative effect of cell transplantation after SCI. However, the present study also sheds light on the risks associated with the use of human iPSCs as appreciable cell survival in the injured spinal cord was associated with a predisposition of grafted cells to form tumors. This was particularly evident in the intact spinal cord. In summary, the present study provides a comparative overview over astrocytic features and pinpoints inclusion or exclusion criteria for transplantation after SCI. These criteria can be used to estimate the pro-regenerative ability of astrocytic populations and to predict the potential of PSC-derived progeny to form tumors.