We differentiated hESCs to the three

We differentiated hESCs to the three primary germ layers in this system, identified extracellular biomarker profiles for each, and confirmed germ layer identity with gene expression analysis. A wide variety of extra- and intracellular markers have been previously utilized to identify each of the primary germ layers. Given the heterogeneous expression patterns of endoderm, mesoderm, and ectoderm, no four-marker combination will perfectly distinguish all pramiracetam within each lineage. CXCR4 is expressed in migrating early endoderm and mesoderm (McGrath et al., 1999), developing vascular endothelial cells (Tachibana et al., 1998), and migrating neural stem cells (Imitola et al., 2004). Although typically 20–25% of mesodermal cells expressed CXCR4, our results showed that endoderm was uniquely distinguishable by CXCR4+KDR– expression at this stage of differentiation. KDR is expressed on mesoderm cells after exiting the primitive streak (Ema et al., 2006) and has been utilized to characterize hESC differentiation to general mesoderm as well as vascular and cardiac progenitors (Nostro et al., 2008; Sakurai et al., 2006; Kattman et al., 2006). Although our results corresponded with previous studies that have found KDR expression in a portion of hESCs (Yang et al., 2008), mesodermal cells generated by high BMP4 conditions were uniquely identifiable by KDR+SSEA-3– expression. Currently, no well-characterized extracellular early ectoderm surface markers have been validated. Ectoderm has typically been characterized by the intracellular markers Nestin, β-III Tubulin, or Sox1 (Vallier et al., 2009; King et al., 2009; Nakagawa et al., 2008; Parekkadan et al., 2008; Smith et al., 2008; Watanabe et al., 2005). We exploited the delayed expression of SSEA-3 into the early neural stem cell stage (Pruszak et al., 2007) to overcome the confounding NCAM expression in endoderm and mesoderm. Transcript expression indicated that this stage is similar to the primitive anterior neuroectoderm/neuroepithelia stage defined by Pankratz et al. (Pankratz et al., 2007), namely that PAX6 expression is upregulated with SOX1 expression beginning to increase. This strategy carries the caveat that SSEA-3+NCAM+ phenotyping would not be applicable at a later stage in the ectoderm lineage due to the loss of SSEA-3 expression. Although exact comparisons are difficult to make due to various time points, markers utilized, data reported, and host species differences, we obtained excellent yields in relation to previous studies. Serum-containing, high activin A protocols have produced SOX17+ endoderm yields ranging from 55% (Borowiak et al., 2009) to greater than 80% (D\'Amour et al., 2005) after 5–6days of differentiation in hESCs. Our yield of 80% for H9 cells falls within the higher range of this spectrum. Day 6 KDR+ mesoderm populations of 15–25% (Goldman et al., 2009) and 52% (Zhang et al., 2008) cells have been obtained from hESCs. We obtained greater than 74% KDR+SSEA-3– mesoderm yields for multiple hESC lines. Patani et al. generated Day 4 and Day 8 yields of 51% and approximately 80%, respectively, for cells positive for the neuroectoderm marker MUSASHI from hESCs (Patani et al., 2009). Chambers et al. generated greater than 80% HES5+ neuroectoderm after 11days of hESC differentiation (Chambers et al., 2009). Our 6-day protocol generated yields of 41–57%, comparable to these studies given the shorter time duration. Many cytokine conditions, small molecules, and transfection conditions may induce cellular toxicity. Additionally, seeding densities may be highly variable due to the inability to seed hESCs as single cell suspensions in many lines. It is vital to obtain consistent and reliable results from a screening assay despite these caveats. This small-scale culture system demonstrated remarkable interwell repeatability and density robustness, particularly since static EB differentiation systems have been reported to generate highly heterogeneous cell populations (Carpenedo et al., 2007). Since 96- and 384-plate assays can display variations due to edges, evaporation, thermal variations, and pipetting error (Armknecht et al., 2005; Barker and Diamond, 2008; Faessel et al., 1999), we utilized gas-permeable membranes and stainless-steel lids to minimize evaporation and ensure a uniform heat distribution (Lucumi et al., 2010). All standard deviations were less than 4% for biomarker signals, which allowed for highly significant differences between conditions.