Supplemental Figure 1:
(A). HSC-3 cells were subjected to MCA formation with or without 10% FBS. After 24 h, MCAs were obtained by brief centrifugation (1000 rpm, 2 min), trypsinized, and analyzed for cell viability by dye exclusion method using trypan blue. (B) MCA obtained in absence of serum was collected as above and allowed to re-attach on tissue culture plates in DMEM media. Photomicrograph of the MCAs at different times show the capability of the MCA to reattach, spread and scatter. Though not provided, MCAs generated in presence of serum also show similar attachment and scattering behavior . Together the results show that serum-deprived MCA are viable similar to MCAs grown in presence of serum for 24 h.
Supplemental Figure 2:
(A). HEK293 cells transfected with the indicated constructs were biotinylated, lysed and immunoprecipitated with anti-Flag, and detected with streptavidin-HRP.
(B). HEK293 cells transfected with vector alone (V) or with ILR-EGFR were treated with varying doses (0, 5, 50, 250, 500, 103 and 106 nM) of EGFR kinase inhibitor, AG1478 for 60 min. Cell lysates were prepared and the phosphorylation activity of the ILR-EGFR was examined by immunoblotting with pEGFR. The membrane was stripped and reprobed with flag-M2 or tubulin as a loading control. Though data not provided there was no effect with SU6656 (C). HEK293 cells transiently expressing the indicated constructs were cultured in the absence of serum for 18 h and with or without EGF (10 ng/ml). Cells were then assayed for DNA synthesis using the CyQUANT cell proliferation kit (Molecular Probes, USA). Briefly, HEK293 cells on 96 well plates were transiently transfected with various constructs. After 24 h post transfection, cells were refreshed with serum free DMEM media and cultured further for 18 h with or without EGF. DNA synthesis was then measured according to the manufacturer’s instructions. The relative fluorescence unit reflects the content of DNA that correlates to the rate of cell proliferation.. Data points are the mean ± SD from 5 replicates (* p<0.05 compared to vector control).
(D). HEK293 cells transfected as in (A) were detached and cultured as single cells in methylcellulose medium without serum. Cells expressing full-length wild type EGFR were also stimulated with EGF (5 ng/ml). After 24 h, cells were collected by centrifugation, washed in PBS, and fixed in 50 mM glycine-HCl, pH 2.0. Cells suspended in PBS were then stained with Hoechst dye 33342 (Sigma, Inc., 0.5 mg/ml), spread on glass slides, and examined under a Zeiss Axiovert 200M microscope. Cells showing two or more nuclear fragments (white arrows) were scored as cells undergoing apoptosis. Twenty to fifty cells from 5-8 random microscopic fields for each experimental set were counted and cell death was represented as the percent apoptotic cells (right panel). Data are expressed as mean ±SD (* p< 0.05, ** p< 0.001).
Supplemental Figure 3:
(A). The indicated HNSCC cells were subjected to MCA formation without serum. After 20 h, aggregates (pre-formed MCA) were obtained by brief centrifugation (1000 rpm, 2 min) and incubated in 10% FBS containing DMEM on pre-coated poly-HEMA dishes. After 6 hr, MCAs were collected, lysed, and processed for immunoblotting for cyclin D1.
(B) MOK-101 and Ca9-22 cells were selected for EGFR inhibition assay since the two cells responded more efficiently to serum for cyclin D induction within 6 hr. Pre-formed MCAs as above were incubated with serum in the presence of DMSO alone, or 10 mM of AG1478 for 6 h. Cell lysates were prepared and immunoblotted for cyclin D1 and p-EGFR. p-EGFR membrane was stripped and re-probed for total EGFR content. Tubulin was used as an additional protein loading control.
Supplemental Figure 4:
(A). Monolayer HSC3 cells serum-starved overnight were washed once with PBS and incubated with BSA- or hE-cad/Fc-coated beads in DMEM. After 4 h, unattached beads were removed by gently rinsing once with PBS, followed by fixation in 3% paraformaldehyde. Cells were then incubated with anti-E-cadherin (She-78, mab) and anti-EGFR (pab) for immunofluorescence evaluation. White arrowheads indicate cell-hE-cad/Fc-beads contact points showing EGFR recruitment. Note that the hE-cad/Fc-beads reacted strongly with the monoclonal antibody that recognizes the extracellular E-cadherin epitope. This was not observed with BSA coated beads. (B). HSC3 cells were detached in PBS containing 4 mM EDTA and 0.05% trypsin. After passing through a cell strainer, cells were subjected to MCA culture in DMEM in the presence of beads pre-coated with BSA or hE-cad/Fc for 4 h. Top panel: MCA with BSA-beads; bottom panel: MCA with hE-cad/Fc-beads. The dark arrowheads indicate beads attached to the MCAs.
(C). HEK293 cells (lanes 1), or cells stably expressing wild type EGFR alone (lanes 2), or co-expressing E-cadherin (Ec1M, lanes 3 and Ec1DM, lanes 4) were non-enzymatically detached and cultured in serum free medium for time durations of 0, 30, 60, and 120 min. Cells were collected and lysates were prepared for Western blotting analysis as indicated. No EGFR phosphorylation was detected in normal HEK293 (lanes 1), but a basal level was evident in HEK293-EGFR cells at 60 and 120 min post-seeding (lanes 2). However, HEK293-EGFR cells co-expressing the Ec1M or Ec1∆M displayed strong phosphorylation as early as 30 min under the same experimental conditions (compare lanes 3 and 4 to lane 2). We observed that the EGFR phosphorylation persisted beyond 4 h culture. These results indicate that E-cadherin, independent of b-catenin and p120 binding, is capable of transactivating EGFR phosphorylation in HEK293 cells. Note: the upper band of E-cadherin has been suggested to be a precursor protein [Chitaev NA, Troyanovsky SM (1998). J Cell Biol 142: 837-46]
Supplemental Figure 5:
(A). Expression analysis of GFP-AKT: Monolayer HSC-3 cells were infected with adenovirus for GFP, KD-AKT (GFP-tagged kinase dead AKT), and CA-AKT (GFP-tagged constitutively active AKT) for 18 h. Infected cells were detached and subjected to MCA formation in DMEM for 20 h. The GFP-protein expressing cells in MCAs were visualized using a Zeiss Axiovert 200M fluorescence microscope with an AxioCam camera (Zeiss, Jena, Germany).
(B). MCAs generated as in (A) were incubated further for 6 h with or without serum and with 1 mm of PD168393 or U012 as indicated. MCAs were collected, lysed, and analyzed for phospho-AKT. Addition of serum to CA-AKT expressing MCA did not further enhance AKT activity. As expected, inhibitors of EGFR or ERK1/2 did not alter AKT activity. Grb2 was included as total protein loading control.