47 STEM CELLS AND DEVELOPMENT Volume 18, Number 1, 2009 © Mary Ann Liebert, Inc. DOI: 10.1089/scd.2007.0266 Teratoma Formation by Human Embryonic Stem Cells Is Site Dependent and Enhanced by the Presence of Matrigel Tatyana A. Prokhorova,* , † Linda M. Harkness,* Ulrik Frandsen,* Nicholas Ditzel, Henrik D. Schrøder, Jorge S. Burns, and Moustapha Kassem When implanted into immunodeicient mice, human embryonic stem cells (hESCs) give rise to teratoma, tumor- like formations containing tissues belonging to all three germ layers. The ability to form teratoma is a sine qua non characteristic of pluripotent stem cells. However, limited data are available regarding the effects of implantation site and the methods employed for implantation on the success rate of teratoma formation. In this study, the rate of teratoma formation in immunodeicient mice was site dependent: subcutaneous (25–100%), intratesticular (60%), intramuscular (12.5%), and under the kidney capsule (100%). Co-injecting the hESCs with Matrigel increased subcutaneous teratoma formation eficiency from 25–40% to 80–100%. We did not observe site-speciic differences in the teratoma composition at the histological level. However, subcutaneous teratomas were quite distinct, easy to remove, and caused minimal discomfort to the mice. Also, subcutaneous teratomas displayed larger proportion of solid tissues as opposed to cyst formation that dominated the teratomas formed at the other sites. Interestingly, a chromosomally abnormal hESCs with trisomy 20 formed teratomas where the ratio of differentiated to undifferentiated tissues was signiicantly decreased suggesting defective pluripotency of the cells. In conclusion, subcutaneous implantation of hESCs in presence of Matrigel appears to be the most eficient, reproducible, and the easiest approach for teratoma formation by hESCs. Also, teratoma formation can be employed to study the development defects exhibited by the chromosomally abnormal hESC lines. Introduction P luripotency ( the ability to differentiate under proper induction conditions into cells and tissues of the three germ layers: ectoderm, mesoderm, and endoderm) is the deining feature of human embryonic stem cells (hESCs) (reviewed in ref. 1). Several assays are currently available to test the pluripotency of hESCs which usually involve spon- taneous or directed differentiation of the cells in vitro and in vivo [1]. In vitro pluripotency is veriied by demonstrating the ability of the cells to form 3D structures known as embry- oid bodies (EBs) containing cells belonging to the three germ layers. The ability to form teratoma when implanted into a host animal, commonly mouse, is a classical in vivo assay for demonstrating the pluripotency of hESCs (reviewed in ref. 2). For mouse embryonic stem cells, chimera formation is another powerful assay, and it has been attempted to cre- ate human–mouse [3] and human–chicken [4] chimera using hESCs. However, such experiments cannot be performed with hESCs on a routine basis due to ethical considerations and therefore teratoma formation in mice remains the stan- dard assay for hESCs pluripotency. Teratomas are nonmalignant tumors that develop in mice from transplanted embryonic stem (or embryoid carcinoma) cells [2,5]. In spite of its importance as a screen for the pluripo- tency of hESCs, there is no standard procedure for perform- ing this assay. Various methods have been described in the literature where hESCs have been implanted under the kid- ney capsule [6], intramuscularly [7], intratesticularly [8], and more recently subcutaneously [9]. In addition, it is not clear whether there exist site-speciic differences in the eficiency of teratoma formation or in the tissue composition. It is plau- sible that a speciic implantation site can affect the degree and the type of differentiation in the teratoma due to the inlu- ence of microenvironment for hESC differentiation. However, Molecular Endocrinology Laboratory (KMEB), Department of Endocrinology and Metabolism, University Hospital of Odense and Medical Biotechnology Center, University of Southern Denmark, Winsløwparken, Odense, Denmark. * These authors contributed equally to this work. † Current afiliation: Center for Eksperimentel Bionformatik (CEBI), University of Southern Denmark, Odense, Denmark.