HBC-3 cells were maintained as previously described (Rogler,L.E.,Am.J. Path. 150:591-602). Cells grown for microarray analysis were stripped of their STON+ feeder layer by trypsinization and replating on plastic in HBM Undifferentiated cells were harvested the next A.M. for preparation of total RNA. Twenty 10cm plates were used to obtain sufficient quantities of undifferentiated reference total RNA. Total RNAs were prepared using Qiagen RNeasy maxi kit. The quality of each RNA preparation was determined using an Agilent 2100 Bioanalyzer. Five microgram aliquots of total RNA from undifferentiated HBC-3 cells were used to generate amplified RNA (aRNA) using the Ambion Message Amp Kit (cat.# 1750). The quality of the aRNA was determined using an Agilent Bioanalyzer. Amplified RNAs were stored as 5ug aliquots under 100% Ethanol at -80oC. Cy 3 and Cy5 labeled targets were synthesized using a modified version of the AECOM functional Genomics facility protocol (http://microarray1k.aecom.yu.edu/) as previously described (Plescia et al, 2001,Differentiation 68:254-269). Cy3 and Cy5 labeled targets were prepared using 5 ug of aRNA using random primers and Superscript reverse transcriptase (Invitrogen) Cy3 labeled target prepared from undifferentiated HBC-3 cells provided the reference sample for all comparisons. The precipitated targets were then resuspended in hybridization buffer as previously described and applied to a microarray. Arrays were hybridized at 50oC for 20 hours. Washing was carried out as described at (http://microarray1k.aecom.yu.edu/ Cy3 and Cy5 labeled targets were synthesized using a modified version of the AECOM functional Genomics facility protocol (http://microarray1k.aecom.yu.edu/) as previously described (Plescia et al, 2001,Differentiation 68:254-269). Cy3 and Cy5 labeled targets were prepared using 5 ug of aRNA using random primers and Superscript reverse transcriptase (Invitrogen).). The purified Cy5 labeled targets were repared from aRNA made from HBC-3 cells treated with Matrigel for 5 day (to induce bile duct differentiation).Cy3 reference target and a Cy5 target prepared from an RNA isolated from cells at one of the time points of Matrigel treatment were combined and precipitated. The precipitated targets were then resuspended in hybridization buffer as previously described and applied to a microarray. Arrays were hybridized at 50oC for 20 hours. Washing was carried out as described at (http://microarray1k.aecom.yu.edu/). " In vitro bile ductular differentiation of HBC-3 cells was performed as previously described (Rogler,L.E.,Am.J. Path. 150:591-602) Cells grown for microarray analysis were stripped of their STON+ feeder layer by trypsinization and on a 1mm Matrigel cushion (differentiated). Cells undergoing Matrigel induced differentiation were harvested at 1 day of Matrigel treatment for preparation of total RNA. Ten 10 cm plates were harvested for each time point to obtain total RNA sufficient for several target preparations at each time point. Five microgram aliquots of total RNA from undifferentiated HBC-3 or Matrigel treated HBC-3 cells were used to generate amplified RNA (aRNA) using the Ambion Message Amp Kit (cat.# 1750). The quality of the aRNA was determined using an Agilent Bioanalyzer. Amplified RNAs were stored as 5ug aliquots under 100% Ethanol at -80oC. Following hybridization and washing, the arrays were scanned for Cy3 and y5 florescence using a scanner Genepix 3.0 scanner (Axon Instruments Inc., Union City, CA). The signal intensity of a spot was considered to be significant if the signal intensity was greater by 200 units of the background for that spot and if it was not flagged as ""not found"" by GenePix software. These data were normalized using the Lowess implementation included in BASE. Following filtration and normalization, the ratio of red and green fluorescence, (RAT) was quantified for each qualified gene. The RAT values were then Log transformed and the LOG2 of the RAT is presented in the data table." Keywords = HBC-3 cells, murine hepatoblast cells, matrigel