Commonly Used Pancreatic Stellate Cell Cultures Differ Phenotypically and in Their Interactions with Pancreatic Cancer Cells

Abstract

Activated pancreatic stellate cells (PSCs) play a central role in the tumor stroma of pancreatic ductal adenocarcinoma (PDAC). Given the limited availability of patient-derived PSCs from PDAC, immortalized PSC cell lines of murine and human origin have been established; however, it is not elucidated whether differences in species, organ disease status, donor age, and immortalization alter the PSC phenotype and behavior compared to that of patient-derived primary PSC cultures. Therefore, a panel of commonly used PSC cultures was examined for important phenotypical and functional features: three primary cultures from human PDAC, one primary from normal human pancreas, and three immortalized (one from human, two from murine pancreas). Growth rate was considerably lower in primary PSCs from human PDAC. Basal collagen synthesis varied between the PSC cultures, and TGF-β stimulation increased collagen synthesis only in non-immortalized cultures. Differences in secretome composition were observed along with a divergence in the DNA synthesis, migration, and response to gemcitabine of PDAC cell lines that were grown in conditioned medium from the various PSC cultures. The findings reveal considerable differences in features and functions that are key to PSCs and in the interactions with PDAC. These observations may be relevant to researchers when selecting the most appropriate PSC culture for their experiments.

Keywords: TGF-β; pancreatic cancer; pancreatic stellate cells; tumor-stroma interaction.

Figure 1

Phenotypic characterization of pancreatic stellate cells. (A) For morphological analysis, cells were stained with hematoxylin and eosin (H&E), BODIPY for detection of cytoplasmic lipid droplets and immunostained with anti-α-SMA (green) and anti-vimentin (red) antibodies. Nuclei were stained with DAPI (blue). Scale bar = 100 µM. (B) Cell size of the various PSC cultures was determined by measurement of the area of 50 cells for each PSC culture using FIJI software. (C) Number of positive cells for α-SMA, vimentin, and BODIPY in percentage. (D) Cells were lysed and proteins subjected to immunoblotting using anti-α-SMA and anti-vimentin antibodies. Due to high exposure time required for detection, α-SMA expression in HPaSteC cells is also presented in a separate blot. GAPDH was used as a loading control. PSC, pancreatic stellate cell; hPSC, human primary PDAC-derived PSC culture; HPaSteC, PSCs from normal human pancreas; i-hPSC, immortalized human PSCs; i-mPSC C2 and C3, immortalized mouse PSCs clone 2 and 3.

Figure 2

Growth curves and doubling time. Growth curves (A) and doubling time (B) were determined by counting cell numbers every 24 h for 5 days. (C) Cell proliferation rate was obtained by measuring the percentage of cell viability by the MTT assay at 24 and 72 h after cell seeding. (D) Percentage of cell viability of HPaSteC cells measured by the MTT assay at 24 and 72 h after cell seeding in special or DMEM medium. (E,F) β-galactosidase staining of primary cultures, including hPSCs and HPaSteC, performed at 24 h (day 1) and 120 h (day 5) after cell seeding (E) and percentage positive cells, using ImageJ software for cell counting (F). Data are mean ± SEM of triplicate determinations. ** p < 0.01 comparing average of hPSCs with HPaSteC, i-hPSC, and i-mPSCs for (C). PSC, pancreatic stellate cell; hPSC, human primary PDAC-derived PSC culture; HPaSteC, PSCs from normal human pancreas; i-hPSC, immortalized human PSCs; i-mPSC C2 and C3, immortalized mouse PSCs clone 2 and 3.

Figure 3

Collagen synthesis. (A) Basal collagen synthesis. (B) Collagen synthesis after TGF-β (10 nM) stimulation for 24 h. Measurement based on incorporation of [³H]-proline into collagen. Data are mean ± SEM of triplicate determinations. (C) Cells treated with or without TGF-β for 24 h were lysed and proteins subjected to immunoblotting using antibodies against TGF-β receptor I and II, phospho-Smad-2, Smad-2, phospho-Smad-3, and Smad-3. GAPDH was used as loading control. * p < 0.05, ** p < 0.01 comparing average of hPSCs with HPaSteC, i-mPSCs for (A); ** p < 0.01 comparing control (non-treated) cells with TGF-β treated cells for (B). PSC, pancreatic stellate cell; hPSC, human primary PDAC-derived PSC culture; HPaSteC, PSCs from normal human pancreas; i-hPSC, immortalized human PSCs; i-mPSC C2 and C3, immortalized mouse PSCs clone 2 and 3.

Figure 4

Effect of PSC-CM on pancreatic cancer cell proliferation and migration. (A) Cancer cell proliferation: BxPC-3 and MIA PaCa-2 cells were incubated with SFM or PSC-CMs for 24 h, and DNA synthesis was determined by [³H]-thymidine incorporation assay. Data are mean ± SEM of triplicate determinations. (B) Cancer cell migration: BxPC-3 and MIA PaCa-2 cells were cultured to confluence, and scratch wounds were established. Images of the wound area were taken immediately after the scratches and 24 h after incubation with SFM or PSC-CMs. The wound area was measured using FIJI software. Data are mean ± SEM of eight scratches for each PSC-CM. # p < 0.1, * p < 0.05, ** p < 0.01 comparing control (SFM) and PSC-CM. PSC, pancreatic stellate cell; hPSC, human primary PDAC-derived PSC culture; HPaSteC, PSCs from normal human pancreas; i-hPSC, immortalized human PSCs; i-mPSC C2 and C3, immortalized mouse PSCs clone 2 and 3; PSC-CM, PSC-conditioned medium; SFM, serum-free DMEM.

Figure 5

Effect of PSC-CM on chemosensitivity for gemcitabine of pancreatic cancer cells. (A) Gemcitabine dose response curves: BxPC-3 and MIA PaCa-2 cells (3000 cells/well) seeded on 96-well plates were incubated with varying concentrations of gemcitabine for 48 h and evaluated for cell viability using the MTT assay. IC₅₀ values for gemcitabine were calculated using GraphPad Prism 4.0 software. (B) BxPC-3 and MIA PaCa-2 cells were incubated with PSC-CM for 24 h prior to incubation with gemcitabine (10 μM) for 48 h. Cell viability was determined using the MTT assay. Data are mean ± SEM of triplicate determinations. The table indicates gemcitabine-induced cytotoxicity in percentage and PSC-CM-induced resistance to gemcitabine, calculated by relative reduction in cytotoxicity between SFM and PSC-CM. * p < 0.05, ** p < 0.01 comparing SFM with PSC-CM. PSC, pancreatic stellate cell; hPSC, human primary PDAC-derived PSC culture; HPaSteC, PSCs from normal human pancreas; i-hPSC, immortalized human PSCs; i-mPSC C2 and C3, immortalized mouse PSCs clone 2 and 3; PSC-CM, PSC-conditioned medium; SFM, serum-free DMEM.

Figure 6

Secretome analysis of PSC-CM. Conditioned medium from various PSC cultures were subjected to proteomics analysis using LC-MS/MS. (A) Protein identifications from three replicates for each PSC culture and number of differentially expressed genes. (B,D) Compared to hPSC, differentially secreted proteins by HPaSteC (B) and by i-hPSC (D) detected by LC-MS were interrogated in terms of functional annotation by the DAVID Bioinformatics Resource tool. The representative GO terms cluster groups with top 10 enrichment score are presented. The horizontal axis represents the significance (p-value) for each term, while the vertical axis represents the GO categories. (C,E) Heatmap of protein abundance pattern for the 50 most significantly downregulated and upregulated proteins (fold change). Red and green color indicates high and low expression, respectively. GO, Gene ontology; PSC, pancreatic stellate cell; hPSC, human primary PDAC-derived PSC culture; HPaSteC, PSCs from normal human pancreas; i-hPSC, immortalized human PSCs; PSC-CM, PSC-conditioned medium.