Figure 1.
Overview of experimental design.
HeLa and HeLa/CDDP cells were grown on isotopically light or heavy media. Light and heavy cell lysates from HeLa and HeLa/CDDP were combined to generate forward and reverse mixed SILAC samples. Light and heavy cell lysates of the same cell type were combined to generate control samples. Protein samples were reduced, alkylated and digested with trypsin into peptide samples that were then analyzed by LC-MS/MS. Quantification of SILAC peptide pairs was performed with the software tool, SILACtor. A heat map was used to visualize proteins displaying increased and decreased expression between HeLa and Hela/CDDP. Finally biological network analysis was combined with protein expression level data to identify biological pathways relevant to cisplatin resistance.
Figure 2.
Distribution and heat map analysis of quantified proteins.
A Histograms illustrating of the distribution of RIA’s observed for the control samples (blue) and the mixed samples (yellow). B Histogram of the distribution of averaged RIA’s with overlaid distributions of the proteins identified by ANOVA with significantly decreased ratios (green) and significantly increased ratios (red). Bars are included indicating a±1.25 fold change cutoff which was determined from cumulative distribution analysis of the control sample at the 99% confidence level. C Heat map generated from SILAC data comparing cisplatin resistant HeLa cells to normal HeLa cells. Lane 1 is light cisplatin resistant and heavy normal cells. Lane 2 is heavy cisplatin resistant and light normal cells. Lane 3 is a one to one mixture of light and heavy cisplatin resistant cells. Lane 4 is a one to one mixture of light and heavy normal cells.
Table 1.
Proteins with largest increased and decreased expression levels in HeLa/CDDP.
Figure 3.
Cropped image of Western blot results comparing levels of four proteins (DDB1, Ku80, CD44, DJ-1) between HeLa and HeLa/CDDP. ACTB was also monitored as a quantitative control. * The normalized ratio for each protein was calculated by averaging the signal intensity from three biological replicates from HeLa/CDDP and dividing it by the average signal intensity from 3 biological replicates from HeLa. This ratio was then normalized using the average ratio of signal intensity of ACTB to the respective protein of interest band. In general there is good agreement between the ratios obtained from Western blotting and from SILAC. Full Western blot images are available in Figure S2.
Figure 4.
Expression level and protein interaction network analysis.
Protein interaction network generated with STRING 8.3 [56] and visualized with Cytoscape [57] consisting of 803 proteins connected by 1174 protein-protein interactions. The 134 proteins with significantly increased expression levels (p<0.01) of at least 1.25 fold in cisplatin resistant HeLa cells are shown in red. Major clusters of interacting proteins include those involved in metabolic energy production, protein folding, and cellular signaling. The 147 proteins with significantly decreased expression levels (p<0.01) of at least 1.25 fold in cisplatin resistant HeLa cells are shown in green. Major clusters of interacting proteins include the 40S and 60S ribosomal proteins and ribonucleoproteins. Interactive Cytoscape networks are included as Dataset S1.
Figure 5.
Biological pathway and molecular function networks.
BiNGO [14] generated biological pathway and molecular function networks for proteins with significantly altered expression levels associated with cisplatin resistance. Node size is related to the number of proteins associated with a GO term, while the color relates to the p-value for the statistical significance of the enrichment of a GO term. Interactive Cytoscape networks are included as Dataset S1.