Fig 1.
Phenotypic characterization of MDA-MB-231 isogenic cell lines.
Phase-contrast images of the parental-231, primary tumor (1° tumor)-231, lung-231, and lymph node-231 cell lines are shown in the top two rows. The top row images were photographed using a X10 objective coupled with a X4 phase-contrast ring while the second-row images were duplicate images obtained with a X10 objective coupled with a X10 phase-contrast ring. The optical configuration of the top row gave 3D images. The black scale bars = 50 μM. Growth curves of each cell line are presented below the images of the corresponding cell line with the mean growth-rate given in the bottom right-hand corner of the curves.
Fig 2.
Principle component analyses (PCAs) and hierarchical clustering of all cell lines.
(A) Proteomic-based PCA plots of 231 isogenic cell lines (PC-1, PC-2, and PC-3 represented 44.1, 37.0, and 18.9% of the respectively) and (B) 435 isogenic cell lines (PC-1, PC-2, and PC-3 represented 29.0, 25.7, and 21.5% of the respectively). (C) Proteomic-based hierarchal clustering (heat map) of six 435 isogenic cell lines along with four 231 isogenic cell lines. (D) Transcriptomic-based hierarchal clustering of all cell lines. All analyses indicated that each cell line had distinct proteomic/transcriptomic signatures, which resulted in the cell lines’ clustering into separate groups/clades. As shown beneath the heat maps, proteins (C) or transcripts (D) distributed across rows have been presented as gradations of color from dark blue-to-dark pink, i.e., relative minimal-to-maximal expression levels. Thus, each row of proteins (C) or transcripts (D) has been placed the left of the cell line designations and the associated trees is at the right.
Fig 3.
Cellular pathway hierarchical clustering’s of isogenic metastatic cell lines.
(A) Proteomic-based up- and down-regulated pathway clustering’s. (B) Transcriptome-based up- and down-regulated pathway clustering’s. As shown beneath the heat maps (colored bar), pathways were distributed across rows and shown as gradations of color from dark blue-to-dark pink, i.e., relative minimal-to-maximal expression levels. Each row of pathways is to the left of the cell line designations and the associated tree is at the right.
Table 1.
Proteomic-based unique pathways for the metastatic brain-435 cell line.
Table 2.
Proteomic-based unique pathways for the metastatic liver-435 cell line.
Table 3.
Proteomic-based unique pathways for the metastatic lung-435 cell line.
Table 4.
Proteomic-based unique pathways for the metastatic spine-435 cell line.
Table 5.
Proteomic-based unique pathways for the metastatic lung-231 cell line.
Table 6.
Proteomic-based unique pathways for the metastatic lymph node-231 cell line.
Fig 4.
The up- and down-regulated proteomic-based interconnected network maps of pathways unique to the brain-435 cell line.
The size range of the nodes correlates to the size of the protein sets while the range of hues of the nodes correlates with the q-values, which is correlated to the size of the number of observed proteins. The edges represent the overlap of shared proteins between the connected nodes with the width of the edges representative of the size of the overlap and their color denoting the number of the observed proteins that are shared.
Fig 5.
The up- and down-regulated proteomic-based interconnected pathway network maps of unique to the lung-231 cell line.
The size range of the nodes correlates to the size of the protein sets while the range of hues of the nodes correlates with the q-values, which is correlated to the size of the number of observed proteins. The edges represent the overlap of shared proteins between the connected nodes with the width of the edges representative of the size of the overlap and their color denoting the number of the observed proteins that are shared.
Fig 6.
The up- and down-regulated transcriptomic-based interconnected pathway network maps of unique to the brain-435 cell line.
The size range of the nodes correlates to the size of the transcript (gene) sets while the range of hues of the nodes correlates with the q-values, which is correlated to the size of the number of observed transcripts. The edges represent the overlap of shared transcripts of the connected nodes with the width of the edges representative of the size of the overlap and their color denoting the number of the observed transcripts that are shared.
Fig 7.
The up- and down-regulated transcriptomic-based interconnected pathway network maps of unique to the lung-231 cell line.
The size range of the nodes correlates to the size of the transcript (gene) sets while the range of hues of the nodes correlates with the q-values, which is correlated to the size of the number of observed transcripts. The edges represent the overlap of shared transcripts of the connected nodes with the width of the edges representative of the size of the overlap and their color denoting the number of the observed transcripts that are shared.
Table 7.
Unique pathways found to be common in both proteomic and transcriptomic analyses for the metastatic brain-435 cell line.
Table 8.
Unique pathways found to be common in both proteomic and transcriptomic analyses for the metastatic liver-435 cell line.
Table 9.
Unique pathways found to be common in both proteomic and transcriptomic analyses for the metastatic lung-435 cell line.
Table 10.
Unique pathways found to be common in both proteomic and transcriptomic analyses for the metastatic spine-435 cell line.
Table 11.
Unique pathways found to be common in both proteomic and transcriptomic analyses for the metastatic lung-231 cell line.
Table 12.
Unique pathways found to be common in both proteomic and transcriptomic analyses for the metastatic lymph node-231 cell line.
Fig 8.
Bar plot presentations of quantitative real-time PCR (qRT-PCR) results.
(A) qRT-PCR results for isogenic metastatic 231 cell lines (lung: black bars and lymph node: white bars). (B) qRT-PCR results for isogenic metastatic 435 cell lines (brain: black bars, liver: light gray bars, lung: dark gray bars, and spine: white bars). The genes (x-axis) are plotted against their log2 fold changes (y-axis).
Table 13.
Genes common to proteomic and transcriptomic data sets and their linear fold change (FC) relative to their 1° tumors were verified with qRT-PCR.
Table 14.
Mean IC50 values for the drugs tested against each isogenic cell line.
Table 15.
Summary of linear fold change of IC50 values for metastatic isogenic cell lines relative to their 1° tumors.
Fig 9.
Representative survival plots of triple negative breast cancer patient data (n = 255).
Genes were derived from proteomic-based up regulated pathways (Table 16) that correlated with TNBC patient relapse-free survival (RFS) datasets (Reference: PMID: 20020197). Hazard ratios indicated that high expression (red) of both FLNB and H1F0 significantly correlated with poor RFS.
Table 16.
Proteomic-based up-regulated pathway proteins correlated to human patient survival.