Self-formation of optic cups and storable stratified neural retina from human ESCs.
Yoshiki Sasai, M.D., Ph.D. On August 5, the stem cell community was shocked by news that one of its most admired and respected leaders, Yoshiki Sasai, had died by his own hand. The humiliating circumstances surrounding the retraction of two papers, which he coauthored, proved overwhelming to him. On this terribly sad occasion it is fitting to reflect on the spectacular scientific legacy he leaves behind. Yoshiki Sasai was born in Hyogo, Japan. He received both an M.D. and Ph.D. from Kyoto University School of Medicine. After completing an internship in internal medicine, he studied molecular biology, first in the laboratory of Dr. Shigetada Nakanishi at Kyoto University, and then with Dr. Eddy De Robertis at UCLA. Like many stem cell scientists, Yoshiki began his career as a developmental biologist and came to the stem cell field through his interest in the molecular steps that determine early neural commitment. It was during his postdoctoral work at UCLA that Yoshiki discovered chordin, a gene that, together with noggin and follistatin, produces the key inductive signals secreted by the Spemann organizer during early development of the amphibian embryo. He further demonstrated that chordin promotes neural fate by inhibiting signals that would drive alternative fates—in effect a default model of neural induction. The notion that neural tissues have an intrinsic ability to differentiate, and that under the right permissive conditions only a small nudge might be sufficient to initiate a developmental program, emerged as a concept that would shape the rest of his career. In 1996 he moved to Kyoto University as an Associate Professor, and he advanced to full Professor 2 years later. He joined the RIKEN Center for Developmental Biology at Kobe in 2000 and became Director of the Organogenesis and Neurogenesis Group. During this time he continued to work on the molecular components of neural induction, as well as the signals underlying the genesis and differentiation of neurons. He realized that pluripotent stem cells could be powerful tools for bridging the gap between the study of embryology and applications in regenerative medicine, and he began directing his research toward translational goals. He showed that exposure to stromal-cell-derived inducing activity can produce midbrain dopamine neurons from mouse embryonic stem cells (Kawasaki et al., 2000). Yoshiki also developed protocols for producing a variety of neuronal types, including ventral forebrain inhibitory neurons. Yet, Yoshiki had a vision beyond producing single cell types: he was interested in recapitulating organogenesis with all of its multicellular complexity and intricate tissue organization. Yoshiki had an unmatched ability to decipher the embryo—specifically, to uncover how this developmental marvel generates the extraordinary diversity of cell types that become organized into unique structures, like the pituitary gland, the brain, or the eye. He was able to apply developmental insights to identify the key molecular steps needed to produce specific cell types and self-assembling structures from embryonic stem cells. In a series of stunning papers, Yoshiki demonstrated how complex, 3D organlike tissues, ‘‘organoids,’’ can be produced from mouse or human pluripotent stem cells. As an example, using this approach he and his group were able to mimic the normal development of the pituitary gland (Suga et al., 2011). The pituitary is derived from the oral ectoderm (Rathkes’ pouch), which interacts with the base of the hypothalamus to generate this tiny gland. The pituitary controls many other glands in the body, including the adrenals, thyroid, and gonads. Yoshiki’s laboratory used embryonic stem cells to generate tissue resembling Rathkes’ pouch that consisted of multiple pituitary cell types, including hormone-secreting cells capable of restoring glucocorticoid levels in mice lacking pituitary glands. In 2008, Yoshiki was able to coax embryonic stem cells grown in 3D cultures to form an embryonic neocortex (Eiraku et al., 2008). Like the pituitary cells, cortical cells selforganized into layers that included the ventricular zone, cortical plate, and Cajal Ratzius cells. The resulting projection neurons, when transplanted into the cortex of