Ancer cell lines were selected for our study. All three lines have an epithelial origin and their tumor biology has been well studied in the literature. SW480 was derived from a MedChemExpress Genz 99067 primary adenocarcinoma of the colon from a patient that subsequently relapsed with wide-spread mesenteric lymph nodes metastases that were used to derive the SW620 cell line [7]. The use of a patient-matched set of cell lines buy EAI045 reduces genetic variability and allows for a more controlled comparison of the molecular changes following metastatic progression [8,9]. Multiplexing of all three samples for simultaneous labeling and analysis was achieved through fluorescent cell barcoding. In this technique, cells are labeled with a distinguishing intracellular dye and then pooled together prior to antibody labeling. The identity of each cell line is recognized on the flow cytometer on the basis of fluorescence from either the violet (Horizon Proliferation Dye; VPD450) or blue (carboxyfluorescein succinimidyl ester; CFSE) excitation lasers, while the red laser is reserved for detection of Alexa647 on the secondary antibodies. The SW480 cell line was barcoded by labeling with VPD450 while SW620 cells were unlabeled prior to pooling both cell lines into a single admixed population (Figure 1, see Materials and Methods). The third cell line, HCT116, was barcoded using CFSE and also admixed to generate a single pool comprised of the three different cell lines. We then applied the combination of cells onto an antibody array consisting of 242 antibodies and 9 isotype controls individually allocated across three 96-well plates (Figure 1). The antibodies included in the array provide coverage of nearly every cluster of differentiation (CD) molecule and many other common surface antigens. As such, we were able to probe a broad range of surface proteins and generate signatures for each colon cancer cell line. The 16574785 majority of antibodies (158/242) were either completely unreactive or bound less than 5 of the cells as compared to the respective isotype control in all three cell lines and therefore not further investigated for the purposes of this study (full results shown in Figure S1 and Tables S1, S2, and S3). We found 25 antibodies that reacted with the majority (.50 ) of cells in SW480, SW620, and HCT116 cell lines (Table 1 and Figure S2). Many of these antibodies reached near complete (.95 ) labeling of tumor cells. As expected, proteins related to major histocompatibility complex class I (b2-microglobulin, HLA-A/A2/B/C andBioinformatics analysisTo prioritize our list of TAAs as candidate biomarkers, we performed cross-comparisons to the Oncomine collection of gene expression microarray datasets from colon cancer as well as corresponding normal tissue [10]. As such, we were able to assess expression of the proteins identified in our screen as compared to RNA profiles across multiple investigators, patient populations, and experimental platforms. We focused our examination of TAA expression in normal colonic tissue, normal liver, as well as colon adenoma and adenocarcinoma 11967625 [11?4]. We then selected those genes that were at least two-fold upregulated (p,0.05) over normal tissue. This further refined our TAA list to the overexpressed genes CD44, integrin a6 (CD49f), and integrin b2 (CD49b) as the most promising candidates (Figure 2). Notably, the expression of these genes was significantly higher in cancer than in normal liver, suggesting a possible therapeutic window for targeted th.Ancer cell lines were selected for our study. All three lines have an epithelial origin and their tumor biology has been well studied in the literature. SW480 was derived from a primary adenocarcinoma of the colon from a patient that subsequently relapsed with wide-spread mesenteric lymph nodes metastases that were used to derive the SW620 cell line [7]. The use of a patient-matched set of cell lines reduces genetic variability and allows for a more controlled comparison of the molecular changes following metastatic progression [8,9]. Multiplexing of all three samples for simultaneous labeling and analysis was achieved through fluorescent cell barcoding. In this technique, cells are labeled with a distinguishing intracellular dye and then pooled together prior to antibody labeling. The identity of each cell line is recognized on the flow cytometer on the basis of fluorescence from either the violet (Horizon Proliferation Dye; VPD450) or blue (carboxyfluorescein succinimidyl ester; CFSE) excitation lasers, while the red laser is reserved for detection of Alexa647 on the secondary antibodies. The SW480 cell line was barcoded by labeling with VPD450 while SW620 cells were unlabeled prior to pooling both cell lines into a single admixed population (Figure 1, see Materials and Methods). The third cell line, HCT116, was barcoded using CFSE and also admixed to generate a single pool comprised of the three different cell lines. We then applied the combination of cells onto an antibody array consisting of 242 antibodies and 9 isotype controls individually allocated across three 96-well plates (Figure 1). The antibodies included in the array provide coverage of nearly every cluster of differentiation (CD) molecule and many other common surface antigens. As such, we were able to probe a broad range of surface proteins and generate signatures for each colon cancer cell line. The 16574785 majority of antibodies (158/242) were either completely unreactive or bound less than 5 of the cells as compared to the respective isotype control in all three cell lines and therefore not further investigated for the purposes of this study (full results shown in Figure S1 and Tables S1, S2, and S3). We found 25 antibodies that reacted with the majority (.50 ) of cells in SW480, SW620, and HCT116 cell lines (Table 1 and Figure S2). Many of these antibodies reached near complete (.95 ) labeling of tumor cells. As expected, proteins related to major histocompatibility complex class I (b2-microglobulin, HLA-A/A2/B/C andBioinformatics analysisTo prioritize our list of TAAs as candidate biomarkers, we performed cross-comparisons to the Oncomine collection of gene expression microarray datasets from colon cancer as well as corresponding normal tissue [10]. As such, we were able to assess expression of the proteins identified in our screen as compared to RNA profiles across multiple investigators, patient populations, and experimental platforms. We focused our examination of TAA expression in normal colonic tissue, normal liver, as well as colon adenoma and adenocarcinoma 11967625 [11?4]. We then selected those genes that were at least two-fold upregulated (p,0.05) over normal tissue. This further refined our TAA list to the overexpressed genes CD44, integrin a6 (CD49f), and integrin b2 (CD49b) as the most promising candidates (Figure 2). Notably, the expression of these genes was significantly higher in cancer than in normal liver, suggesting a possible therapeutic window for targeted th.