Human embryonic stem cells (hESCs) represent one of the most promising cell sources to produce unlimited
numbers of insulin producing beta-cells for the treatment of type 1 diabetes. We and others have reported
multi-stage protocols for the differentiation of hESCs into insulin producing cells that are capable of effectively
reversing diabetes in rodents. Currently, hESC derived pancreatic progenitor cells are being tested in clinical
trials in patients with diabetes, but for safety reasons, the cells are contained within macroencapsulation
devices and implanted subcutaneously, which may limit graft survival and function. Therefore, we are
proposing to engineer new hESC lines that will not be detectable by the recipient's immune system and can
thus be transplanted without encapsulation or the use of immune-suppressive agents. Moreover, in order to
enhance the safety profile of these cells, we have designed several `safety-switches' whereby the cells may
be selectively eliminated. Firstly, using a regulatable CRISPR/Cas9 system, we will reversibly disrupt the
expression of the beta 2 microglobulin gene, which is a common subunit essential for cell surface expression
of all HLA class I heterodimers. Mismatched HLA is the basis of humoral and cellular alloimmune responses,
and HLA presentation of peptides mediates the autoimmunity towards beta-cells in subjects with type 1
diabetes. Therefore, disruption of HLA class I in hESCs and their differentiated progeny is hypothesized to
make the transplanted cells invisible to the recipient's immune system, while leaving the immune system
fully competent to defend against opportunistic infections or malignances. Importantly, we have designed
the system such that brief exposure to doxycycline will restore expression of HLA class I, thereby removing
the invisibility cloak from the cells and rendering them subject to immune-based elimination. As a second
independent safety measure, we will incorporate in the cells the expression of the clinically validated
inducible herpes simplex virus thymidine kinase (HSV-TK) under the regulatory control of the Nanog
promoter, a pluripotency gene. In this manner, the harmless prodrug ganciclovir is converted to a toxic
metabolite by HSV-TK, thereby killing any contaminating pluripotent cells that remain following cell
differentiation, thus reducing the risk of teratoma formation following transplant. Thirdly, we will incorporate
the expression of the clinically validated inducible caspase-9 protein (iCasp9) into a constitutively active
locus. By itself, the iCasp9 protein remains monomeric and non-toxic, but when cells are exposed to the
drug AP1903, the iCasp9 molecules homodimerize and rapidly induce apoptosis. Thus, the entire graft can
be eliminated if required by treating the recipient with AP1903. Collectively, this multi-layered approach will
enhance the safety profile of the universal donor cells. If successful, this approach will produce cells useful
for optimizing transplant procedures and improving graft survival and function, without the use of constricting
encapsulation technologies or potent immunoregulatory molecules.