Culture of human embryonic stem cells
For the PowerBlot analysis, enzymatically passaged BG01 hESCs were grown as described previously . These conditions were necessary to scale up the culture to generate the milligram amounts of protein lysate required for this analysis. These conditions maintain cell populations that express the appropriate markers of pluripotency and can differentiate to representatives of all three germ layers, but may lead to eventual accumulation of trisomies for chromosomes 12, 17 or X . For the Kinexus assays, BG03 hESCs were maintained in MEF-conditioned medium (MEF-CM) without the accumulation of karyotypic abnormalities as described previously [14, 26].
hESCs were also maintained in a defined medium as indicated. These conditions are described in detail elsewhere (TCS, AJR, submitted). Briefly, the media consisted of DMEM/F12 (Invitrogen), 2% fatty acid-free Cohn's fraction V BSA (Serologicals), 1× nonessential amino acids, 50 U/ml penicillin/streptomycin, 50 μg/ml ascorbic acid, 10 μg/ml bovine transferrin, 0.1 mM β-mecaptoethanol (all from Invitrogen), 1× Trace Elements A, B & C (Mediatech), 10 ng/ml hergulin1β (Peprotech), 10 ng/ml activinA (R&D Systems), 200 ng/ml LR3-IGF1 (JRH Biosciences), and 8 ng/ml FGF2 (R&D Systems). Cultures were passaged using Collagenase IV and plated on growth factor depleted Matrigel (BD Biosciences) diluted 1:200. These cultures were karyotypically normal.
To partially differentiate hESC cultures for immunostaining analysis, karyotypically normal BG01 cells were plated on matrigel and grown for three days in DMEM/F12 containing 10% fetal calf serum (HyClone), 1× nonessential amino acids, 20 mM L-glutamine, 50 U/ml penicillin/streptomycin, and 0.1 mM β-mecaptoethanol.
BG01 hESC lysate was prepared in 10 mM Tris-HCl pH 7.4, 1 mM sodium orthovanadate and 1% SDS, and the PowerBlot assays were performed by BD Biosciences (BD Biosciences). Briefly, 200 μg of protein lysate was loaded in a single, gel-wide well, on a SDS-4–15% gradient polyacrylamide gel. The full PowerBlot screen consisted of five gels, which were blotted and probed with 934 antibodies, and was performed in duplicate with the same cell lysate. The gel dimensions were 130 × 100 × 0.5 mm, and proteins were separated at 150 volts for 1.5 hours, and transferred to an Immobilon-P membrane (Millipore). The membranes were blocked and clamped in a manifold that created 40 lanes across each membrane. A mix of 1 to 8 mouse monoclonal primary antibodies was added to each lane, in dilutions and combinations that had been predetermined to enable accurate identification of well-separated proteins. The predicted sizes of detectable proteins in the blots ranged from 10–540 kD, and the dilutions of the primary antibodies ranged from 1:250 to 1:15,000.
The blots were removed from the manifolds, washed and incubated with goat anti-mouse secondary antibody conjugated to the Alexa680 fluorophore (Molecular Probes). The membranes were scanned using the Odyssey Imaging System (LI-COR). Molecular weight standards were generated by adding a cocktail of antibodies to P190 (190 kD), Adaptin beta (106 kD), STAT-3 (92 kD), PTP1D (72 kD), Mek-2 (46 kD), RACK-1 (36 kD), GRB-2 (24 kD) and Rap2 (21 kD) to lane 40 of gels A-D. Molecular standards for gel E were generated by adding a cocktail of antibodies to Exportin-1/CRM1 (112 kD), MCM (83 kD), Nucleoporin p62 (62 kD), α-tubulin (55 kD), Actin (42 kD), KNP-1/HES1 (28 kD) and NTF2 (15 kD) to lane 16, and antibodies to p190 (190 kD), Hip1R (120 kD), Transportin (101 kD), Calreticulin (60 kD), Arp3 (50 kD), eIF-6 (27 kD) and Rap2 (21 kD) to lane 17.
Bands were detected and raw signal intensity captured automatically using the PDQuest software (Bio-Rad). To normalize the signal intensities, the total raw quantity of each band was divided by the average intensity value of the molecular standards in that image and the normalized values for the duplicate samples were averaged and expressed as normalized intensity units (i.u.). These values represent the relative signal intensity observed for each identified protein band, rather than relative expression levels of different proteins, due to differences in the efficiencies of antibody binding and dilution of the primary antibodies used. Proteins were identified based on the similarity of expected and observed band migration profiles and bands that could not be identified were excluded from the analysis. All identified proteins were verified by visual inspection, and proteins exhibiting a low signal intensity, with an averaged signal of < 1000 i.u., were verified by visual inspection using contrast enhancement in Adobe Photoshop. Bands with > 800 i.u. could typically be observed without additional image enhancement. Microsoft Excel files were generated that contained information on: gel number, lane number, antibody catalogue number (BD Biosciences), protein name, expected size, observed size, repeat 1 i.u. value, repeat 2 i.u. value, averaged i.u. value, antibody dilution, outline of protein function, Entrez gene and SwissProt identification numbers. These tables were used to list expressed proteins (Additional File 1).
Preparation of the BG03 cell lysate and western blotting was performed according to published protocols . Briefly, cell lysate was prepared in 20 mM MOPS pH 7.0, 2 mM EGTA, 5 mM EDTA, 30 mM sodium fluoride, 40 mM β-glycerolphosphate pH 7.2, 20 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 1 mM PMSF, 3 mM benzamidine, 5 μM pepstatin, 10 μM leupeptin, 0.5% nonidet P-40, with the final pH adjusted to 7.2. The Kinexus assays for protein kinases (KPKS-1.2A and B [76 antibodies]), phosphatases (KPPS-1.2 [27 antibodies]) and phosporylated sites in cell signaling molecules (KPSS-3.1 [37 antibodies]) were performed by Kinexus. The Bio-Rad Mini-PROTEAN 3 electrophoresis system was used to separate proteins by SDS-PAGE. For each assay, 250 μg of cell lysate was loaded in a single well spanning the width of the stacking gel, then separated through a 12.5% SDS-Polyacrylamide gel and transferred to a PVDF membrane. A 20-lane manifold was placed over the membrane and a different mixture of up to 3 primary antibodies was added to each well. The combinations of primary antibodies had been predetermined to detect well-separated proteins, avoiding crossreaction to different proteins that co-migrate. The primary antibodies were rabbit and goat polyclonal, and mouse monoclonal antibodies, diluted 1:1000. After incubation with the primary antibodies, the membranes were removed from the manifolds, washed and incubated with a mix of the appropriate secondary antibodies. The secondary antibodies were donkey anti-rabbit (at 1:5000), sheep anti-mouse (at 1:10,000) and bovine anti-goat (at 1:10,000), all conjugated with horse radish peroxidase. The membranes were washed and immunoreactive bands detected by enhanced chemiluminescence (Amersham-Pharmacia) using a FluorS Max Multi-imager (Bio-Rad). Prestained size markers (201.5, 156.8, 106, 79.7, 48.4, 37.8, 23.3, and 18.2 kD) and predetermined human-specific protein migration profiles were used to accurately identify proteins using the Kinexus immuno-reactivity identification system (IRIS) software. Detected proteins were verified by visual inspection.
Immunocytochemistry and staining procedures were as described previously . Briefly, cells were fixed with 4% paraformaldehyde for half an hour, blocked in blocking buffer (5% goat serum, 1% BSA, 0.1% Triton X-100) for 1 hour followed by incubation with the primary antibody at 4°C overnight. Appropriately coupled secondary antibodies (Molecular Probes) were used for single and double labeling. All secondary antibodies were tested for cross reactivity and non-specific immunoreactivity. The following antibodies were used: ABP-280 (1:250, BD Biosciences 610798), CtBP1 (1:1000, BD Biosciences 612042), CtBP2 (1:1000, BD Biosciences 612044), GS-28 (1:2000, BD Biosciences 611184), HDJ-2 (1:100, BD Biosciences 611872), L-Caldesmon (1:2000, BD Biosciences 610660), Rabaptin-5 (1:500, BD Biosciences 611080), phospho-p130 Cas (Tyr165) (1:50, Cell Signaling Technology 4015), phospho-Ras-GAP (pY460) (1:250, BD Biosciences 612736), Ras-GAP (1:250, BD Biosciences 610043), Shc-C (1:1000, BD Biosciences 610642), Oct-4 (Santa Cruz biotechnology, 1:200 SC-8628), TNIK (1:100, BD Biosciences, 612250), p130 Cas (1:100, BD Biosciences, 610272), ERBB2 (1:100, Lab Vision, 9G6.10), ERBB3 (1:100, R&D Systems, MAB348), ZO1 (1:100, Invitrogen, 61–7300), or Occludin (1:100, Invitrogen, 71–1500). Hoechst (Invitrogen) or DAPI (Sigma) were used to identify nuclei, and Triton X-100 was omitted when staining for extracellular antigens (ZO1, occludin, ERBB2/3). Images were captured on an Olympus or Nikon fluorescence microscope.
lllumina data and comparison to proteomic database
Expression levels of proteins detected by the PowerBlot assay were compared to our previous published database of multiple hESC lines examined using the Illumina bead array platform (Liu et al., 2006). Averaged transcript expression signals from the BG01, BG02 and BG03 cell lines were converted to a +/- format, based on the following criteria: A mean transcript detection level of > 5,000 was designated as ++++; 1,000–5,000 as +++; 100–1,000 as ++; 30–100 as +; and signals < 30 was represented as -. In parallel, the protein expression levels were converted to a +/- format based on these criteria: i.u. > 10,000 as ++++; 5,000–10,000 as +++; 1,000–5,000 as ++; 100–1,000 as +. In addition, genes were categorized into the same functional/signaling pathways as per the western blot database.