Characterization of undifferentiated human ES cells and differentiated EBs by antibodies
All monoclonal antibodies were initially selected for their abilities to recognize recombinant proteins in direct ELISAs. A subset were also tested by Western Blot analysis using recombinant proteins and cell lysate to confirm binding to a single epitope. The best clone was later screened for its applications for immunocytochemistry and flow cytometry using various cell lines. Human peripheral blood platelets were used for screening mouse anti-human CD9 antibody. MCF-7 cells were used for screening mouse anti-human E-Cadherin and PODXL (podocalyxin-like) antibodies. MG-63 cells were used for screening mouse anti-human GATA1 (GATA binding protein 1) antibody. Beta-TC6 cells were used for screening for mouse anti-human/mouse PDX-1 (pancreatic duodenal homeobox-1) antibody. NTERA-2 cells were used for screening mouse anti-human Oct3/4 and SOX2 (sex-determining region Y-box 2) antibodies. All polyclonal antibodies were affinity-purified using recombinant proteins and validated by direct ELISAs and Western. Caco-2 cells were used for validation of goat anti-human GATA6 antibody and NTERA-2 cells were used for validation of goat anti-human Nanog and anti-human Oct3/4 antibodies (Summarized in Table 1).
Table 1  Summary list of antibody verification by western blot.  N/A: 1. DPPA5 is still being subcloned. Only Elisa verification is available.
2. The clone for GATA-1 (MAB1779) does not work for Western blot application but is useful for IHC, The clone picked for Western blot analysis does not work for IHC (MAB17791, see data in ).  After antibodies were validated in direct ELISAs, Western blot or cell lines (Fig. 1 and data not shown), they were used to examine the expression of individual molecules in undifferentiated human ES cells and differentiated EBs. When examined by immunohistochemistry, high level of expressions of Oct3/4, SOX2, E-Cadherin, PODXL and Nanog were observed in undifferentiated human ES cells (Fig. 2A, 2B and 2C). DPPA5 (developmental pluripotency associated 5) expression was also observed in undifferentiated human ES cells (data not shown). We noted that a subset of the proteins used were membrane bound proteins. To test if any of the antibodies generated could recognize an extracellular epitope and thus be used for live cell sorting, we repeated staining of live cells as previously described. The CD9, E-Cadherin and PODXL antibodies recognized an extracellular epitope and their ability to select cells by FACS was confirmed (Fig. 3). Minimal or no expressions of Oct3/4, E-Cadherin, PODXL and Nanog were detected in the differentiated EBs (Fig. 2D, 2E and 2F). However, SOX2 expression, which is observed in neural progenitor cells, is persistent in subsets of EBs.
Figure 1  Western blot analysis for Gt × hOct3/4 (A), Gt × hNanog (B) and Ms × hSOX2 (C) in NTERA-2 cell lysate, Ms × hE-Cadherin (D) in MCF-7 cell lysate, Ms × hCD9 (E) in PBMC lysate and Ms × hPDX-1(F) in β-TC-6 cell lysate. Numbers indicate the positions of molecular weight markers.
Figure 2  Undifferentiated human ES cells (A, B, and C) and differentiated EBs (D, E and F) were analyzed using antibodies to indicated molecular markers. Immunostaining with goat anti-human Oct3/4 (Red in A and D), mouse anti-human SOX2 (Green in A and D), goat anti-human E-Cadherin (Red in B and E), mouse anti-human PODXL (Green in B and E), and goat anti-human Nanog (Red in C and F), are contrasted with DAPI nuclear staining (Blue in C-F). Note the dramatic downregulation of ESC specific markers (Oct3/4, E-Cadherin, PODXL, and Nanog) in EBs. However, SOX2 expression is persistent in subsets of EB cells. Scale bars = 100 μm.
Figure 3  Human embryonic stem cells stained with anti-CD9 (A), anti-E-Cadherin (B), and anti-PODXL (C) and antigen expression detected by a flow cytometer. The specific staining is indicated by green histogram and corresponding isotype control is indicated by black histogram.   Suspension culture with FGF withdrawal is known to induce differentiation of ES cells to all three germ layer precursors [12]. The differentiation status of the EB used here was detected to contain all germ cell markers by RT-PCR (Fig. 4). In order to examine how more antibodies can be used for characterization of early differentiation events from human ES cells, we examined the expressions of endodermal markers, SOX17, GATA6 and PDX-1, and mesodermal markers, Brachyury and GATA1, in the undifferentiated human ES cells and differentiated EBs. Expressions of SOX17, GATA6, PDX-1, Brachyury and GATA1 were not detected in undifferentiated human ES cells (data not shown). In contrast to the undifferentiated ES cells, subpopulations of SOX17-, GATA6-, Brachyury- and GATA1-positive cells were observed (Fig 4). These results suggest that both endodermal and mesodermal precursors exist in EBs with FGF withdrawal for 8 days. However, no PDX-1-positive cells were seen in EBs differentiated with the same treatment (data not shown).
Figure 4  Differentiated EBs were analyzed by either immunocytochemistry or RT-PCR to the indicated molecular markers. (A) Immunostaining with goat anti-human SOX17 (Red), is contrasted with Fluoro Nissl nuclear staining (Green). (B) Immunostaining with goat anti-human GATA6 (Red), is contrasted with DAPI nuclear staining (Blue). (C) Immunostaining with goat anti-human brachyury (Red), is contrasted with DAPI nuclear staining (Blue). (D) Immunostaining with mouse anti-human GATA1 (Red). Note that each antibody recognizes subsets of EB cells. Scale bars = 100 μm. (E) The differentiation status of EB is detected by RT-PCR using different germ layer cell markers. Selected endoderm markers AFP, FoxA2; mesoderm markers Hand1, MSX1 and ectoderm marker Msl1 were all highly expressed in the EB samples while their expression was either undetectable or at low level in the ES samples. G3PDH was a positive control showing similar amount of RNA samples were used for analysis.

E