Establishment of two oxaliplatin-resistant gallbladder cancer cell lines and comprehensive analysis of dysregulated genes

2.1 Cells and cell culture

The human GBC cell line GBC-SD was purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) and SGC996 was purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). The two sensitive cell lines were cultured in Dulbecco’s modified eagle’s medium (DMEM; BBI Life Science, Shanghai, China) supplemented with 10% fetal bovine serum (FBS; BI, Israel) and 1% penicillin/streptomycin (BBI Life Science). Cells were maintained in an incubator with 5% CO₂ at 37℃.

2.2 Establishment of OXA-resistant GBC-SD and SGC996 cell lines

The sensitive cell lines GBC-SD and SGC996 were seeded in six-well plates, and resistant cells were induced by gradually increasing the drug concentration. The process can be divided into four stages. Stage I: The initial concentration of OXA (Selleck, Shanghai, China) was 0.05 μM and the passage cycle was controlled at 2–3 days at a seeding density of 2 × 10⁵ cells/well. This stage lasted for 2–3 weeks until the cells were fully adapted to the initial dose. Stage II: The dose was gradually increased until the cells could survive and proliferate in medium containing 5 μM OXA, and this stage lasted for 10–12 weeks. The passage cycle was controlled for 4–5 days at a seeding density of 3 × 10⁵ cells/well. Stage III: The achieved cells were preserved in liquid nitrogen for a month and then recovered and cultured with 5 μM OXA in every passage for an additional month. The passage cycle and seeding density were similar to those of stage II. In stage III, cells underwent an additional cycle of cryopreservation and normal growth with 5 μM OXA to obtain stable OXA-resistant cell lines, referred to as GBC-SD/OXA and SGC996/OXA. Stage IV: OXA (1 μM) was used as the common dose in long-term cryopreservation and passage culture. Cell morphology was observed and photographed using a microscope. Only the resistant cells at stage IV were used for the following assays, including RNA-seq analysis.

2.3 Drug-sensitivity assay

Two sensitive and two resistant cell lines were seeded at 5 × 10³ cells/well in 96-well plates, and different concentrations of OXA (0, 0.01, 0.1, 1, 10, and 100 µM) were added to the wells. Each concentration was used in four replicate wells. After 72 h of culture, the absorbance of each well was measured at 490 nm using an 2,5-diphenyl-2H-tetrazolium bromide (MTT) kit (CT01; Millipore), following the manufacturer’s protocols. The survival rate (%) was calculated according to the following equation: Cell viability (%) = OD of the experimental group/OD of the control group × 100. The drug concentration–viability curve and half-inhibitory concentration (IC₅₀) were calculated using GraphPad Prism 5.0 (GraphPad Software, La Jolla, CA, USA). The drug resistance index (RI) was calculated according to the following equation: RI = IC₅₀ of resistant cell line/IC₅₀ of sensitive cell line.

2.4 Proliferation assay

The four cell lines were seeded at a density of 3 × 10³ cells/well in 96-well plates. The viability of the cells was measured using the MTT kit daily from days 1–7. Four replicate wells were used for each experiment. The proliferation rate was calculated using the equation: Proliferation rate (%) = OD of day n/OD of day 1, (n = 1–7).

2.5 Migration assay

The cells were collected and washed twice with FBS-free DMEM. An 8 μm pore polycarbonate membrane Boyden chamber insert in a Transwell apparatus (Millipore, MA, USA) was used to detect cell motility. In brief, 4 × 10⁴ cells resuspended in 0.2 mL FBS-free DMEM were seeded in the upper chamber, and 0.6 mL of the complete DMEM was added to the lower chamber. After incubation for 24 h, the cells on the upper surface were removed and the migrated cells on the lower surface were stained with 0.1% crystal violet for 10 min. Finally, the number of migrated cells was determined by counting the stained cells in three randomly selected fields using a microscope. The assay for each cell line was performed in triplicates.

2.6 Western blot

Total protein from the four cell lines was lysed using RIPA buffer (Beyotime, Shanghai, China) and quantified using a Bradford protein assay kit (MultiSciences, Hangzhou, China), following the manufacturer’s instructions. Next 10% sodium dodecyl-sulfate polyacrylamide gel electrophoresis was performed, and 10 μg of protein was added to the lanes. After electrophoresis was completed, separated proteins distributed in the gels were transferred into the 0.22 μm PVDF membranes. Subsequently, the membranes were incubated with 5% non-fat milk for 2 h and washed thrice with Tris-buffered saline with 0.1% Tween® 20 (TBST). The membranes were then cut into several pieces following the molecular weight of the target proteins, and incubated with the corresponding primary antibody overnight at 4℃, followed by the incubation with the secondary antibody at room temperature for an additional 2 h. Finally, the bands were incubated with the enhancement buffer following the steps of the ECL kit (Beyotime), and the signals were captured using an imaging system. The primary antibodies, E-cadherin, vimentin, N-cadherin, P-glycoprotein (P-gp), ERK1/2, p-ERK1/2, PD-L1, and the control, GAPDH, were purchased from ABclonal Biotech (Wuhan, China) and diluted 1,000-fold before incubation. The secondary antibody (1:2,000) was purchased from Cell Signaling Technology (Danvers, MA, USA). The gray values of the bands were analyzed using ImageJ software (National Institutes of Health, MD, USA).

2.7 Colony formation assay

For each cell line, 2 mL of complete medium containing 400 cells was added to a six-well plate. After 3 days of culture, 1 μM OXA was added to each well and the cells were cultured with OXA for an additional 11 days. The medium was changed every 4 days. Finally, the supernatant was removed and the cells were stained with 0.1% crystal violet for 20 min. Colonies containing >30 cells were counted and the assay was repeated thrice.

2.8 Soft agar colony formation assay

First, 1.5 and 0.8% agarose solutions were prepared in phosphate-buffered saline and subjected to high-pressure sterilization. Next a 1.5% agarose solution was mixed with an equal volume of medium and pre-coated in a 6-well plate. After the agarose solution solidified, 0.8% agarose solution mixed with an equal volume of medium containing 1,200 cells was seeded into each well. Colony formation lasted for 14 days, and 200 μL of medium was added to each well every 2 days.

2.9 RNA-seq analysis

Total RNA was collected from the four cell lines using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and extracted using an RNeasy mini kit (Qiagen, Germany). RNA quality and concentration were determined using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Montchanin, DE, USA) and assessed using 1% gel electrophoresis. Next the four qualified RNA samples were sent to Sinotech Genomics Co., Ltd (Shanghai, China), for cDNA library construction and RNA sequencing. Gene abundance was expressed as fragments per kilobase of exon per million reads mapped (FPKM). StringTie software was used to count the fragments within each gene, and the TMM algorithm was used for normalization. The DGs between the sensitive and resistant groups were screened under the threshold of |log2 (fold change [FC])| > 1.

2.10 Quantitative polymerase chain reaction (qPCR)

Total RNAs were extracted from the four cell lines using a Universal RNA Extraction Kit (TAKARA, Dalian, China), following the manufacturer’s protocol, and then quantified using the NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies). Next 1 μg of RNA was reverse-transcribed into cDNA using a PrimeScript RT reagent Kit (TaKaRa), according to the manufacturer’s instructions. For quantification of the target genes, the reaction system (20 μL) containing 2 μL of template, 10 μL of TaqMan Universal PCR Master Mix (Applied Biosystems), 1 μL of forward and reverse primers, and 7 μL of water was prepared and subjected to a Bio-Rad CFX96 Thermal Cycler (Bio-Rad, Hercules, USA). The thermal cycling conditions were as follows: 95℃ for 5 min, followed by 40 cycles at 95℃ for 15 s, and 60℃ for 60 s. Relative expression levels of mRNAs were calculated using the 2^−ΔΔCT method, and GAPDH was used as the internal control. The sequences of the primers listed in Table 1 were synthesized by Sangon Biotech (Shanghai, China).

2.11 Bioinformatical analysis

Gene Ontology (GO) analysis for biological processes, cellular components, and molecular function and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis (http://www.genome.ad.jp/kegg) were performed to clarify the potential biological roles of DGs using the enrich R package. Enrichment analysis was conducted to list the top 30 GO terms and KEGG pathways.

2.12 Statistical analysis

Statistical analysis was performed using SPSS software (version 20.0; Chicago, IL, USA), and GraphPad Prism 5.0 (La Jolla, CA, USA) was used for data presentation. Statistical significance was tested using the Student’s t-test. p < 0.05 was considered statistically significant.

3.1 Drug-sensitivity detection of OXA-resistant cell lines

The morphology of GBC-SD and GBC-SD/OXA was fusiform, and the resistant cells were larger than the parental cells (Figure 1a). Sensitive SGC996 cells were small and polygonal, whereas the resistant cells became significant spindles, and some of them had several pseudopodia (Figure 1a). In addition, the two sensitive cell lines tended to aggregate, whereas the resistant cells were both more dispersed (Figure 1a), which may be related to the alteration of motility.

Figure 1

Drug-sensitivity detection of OXA-resistant cell lines. (a) Morphological character of resistant cell lines. Scale bar: 100 μM, magnification: 200×. (b) Cells were incubated with different concentrations of OXA and the MTT assay was used to detect drug-sensitivity.

An MTT assay was used to detect the drug sensitivity of the four cell lines. As shown in Figure 1b, GBC-SD and GBC-SD/OXA cells were similar in response to low and medium doses of OXA (0.01–10 μM), and GBC-SD/OXA showed a significantly higher resistance to 100 μM OXA. The IC₅₀ values of GBC-SD and GBC-SD/OXA were 28.49 and 141.10 μM, respectively, and the RI value was 4.95. SGC996 cells were more sensitive to OXA than SGC996/OXA at in the dose of 1–100 μM. The IC₅₀ value of SGC996/OXA cells was 29.02 μM, which was approximately 4-folds that of the sensitive cells (7.08 μM). These results confirm that the OXA-resistant cell lines GBC-SD/OXA and SGC996/OXA were successfully established.

3.2 Phenotype alterations of OXA-resistant cells

As shown in Figure 2a, the proliferation rates of GBC-SD and its resistant cells were similar in the first 3 days of the study, and GBC-SD/OXA exhibited a slower growth rate than the sensitive cells in the last 3 days. Similar proliferation characteristics were observed in the SGC996/OXA cells (Figure 2a). Clone formation assay results indicated that resistant cells showed a stronger cloning potency than sensitive cells (Figure 2b) in the presence of 1 μM OXA, suggesting better tolerance of the resistant cells. Furthermore, soft agar colony formation assay results also indicated a higher three-dimensional cell sphere formation ability of the two resistant cell lines (Figure 2d), which indicated better stemness after cells were resistant to OXA. In addition, the migration abilities of the two resistant cell lines were significantly enhanced by the Transwell assay (Figure 2d). Western blotting results showed that vimentin and N-cadherin protein expression was both significantly increased after cells were resistant to OXA (Figure 2e). There was no significant change in E-cadherin expression in GBC-SD/OXA and GBC-SD cells, whereas its level was decreased in SGC996/OXA compared to that in the sensitive cells (Figure 2e). Overall, the resistant cells underwent epithelial mesenchymal transformation (EMT), consistent with their enhanced migration ability. One of the classic drug-resistant proteins, P-gp, was also detected, and the results showed a significant increase in the protein levels in the resistant cells (Figure 2e).

Figure 2

Phenotype detection of OXA-resistant cells. (a) Cells were cultured for 1–7 days, and the MTT assay was used to detect the proliferation rate daily. (b) Cells were cultured for 3 days and incubated for an additional 11 days in the presence of 1 μM OXA to evaluate colony formation ability. (c) The soft agar colony formation assay was performed to detect cell sphere formation ability. Scale bar: 100 μM. (d) Transwell assays were performed to detect the migration ability. Scale bar: 100 μM. (e) Western blotting was used to analyze EMT and P-glycoprotein markers.

3.3 RNA-seq analysis of gene expression profile of OXA-resistant cell lines

A total of 4,675 DGs were screened under the threshold of |log2[fold change (FC)]| > 1, including 2,060 upregulated and 2,615 downregulated genes (Figure 3a). The heatmap summarizes the gene expression of all DGs in the four samples, and the colors in each line indicate the FPKM values varying from +∞ to −∞ of the gene in each sample (Figure 3b). The gene expression files of the two different sensitive cells varied significantly, and several of them showed opposite FPKM values (positive and negative numbers). This was one major reason for the p-value not being considered in the screening of DGs. Interestingly, there was a significant difference in the FPKM values between the sensitive and resistant cell lines. The calculation of the final FC value was based on the mean FPKM value of the two sensitive/resistant samples in the same group; thus, thousands of genes were screened although there was no significant alteration in some genes (for example, similar watchet blue of FPKM values) between each pair of the sensitive and resistant cell lines. The DGs identified using RNA-seq provided several candidates for the study of resistance mechanisms, which also required further validation. The FC values of all DGs are listed in Table A1.

Figure 3

RNA-seq analysis of gene expression profile of OXA-resistant cell lines. (a) Scatter plot of DGs in the resistant groups. G2: resistant group, G1: sensitive group. (b) Heat map summarizing the expression of the DGs in each cell line. The color (red-blue) indicates that the FPKM values vary from +∞ to −∞. Each line represents a gene and each column represents a sample or cell line.

3.4 Validation of the DGs by qPCR

Because the alteration trend of FPKM values of some DGs in each pair of cell lines was inconsistent, 12 genes were randomly selected from the top 300 DGs used for qPCR validation. The FPKM and FC values are presented in Table 2. 50% of the FPKM values in SGC996 cells and its resistant cells were 0, and the calculated log2FC value was predominantly contributed by GBC-SD and its resistant cells. The qPCR validation results indicated that the expression levels of most of the selected DGs were significantly changed in each resistant cell line (Figure 4). Among the 12 randomly selected genes, the expression changes of the 9 genes, accounting for 75%, in GBC-SD/OXA were consistent with the RNA-seq results. For SGC996/OXA, the alteration trends of five genes (CSF2, S1PR1, DLX1, MED29, and CXCL8) were conforming to the final log2FC values of RNA-seq (Figure 4, Table 2). In addition, the mRNA levels of above five genes were also both significantly enhanced in the two resistant cell lines (Figure 4). Overall, the RNA-seq results were reliable and satisfactory through the validation of random DGs, although some FPKM values were zero.

Figure 4

qPCR validation of the mRNA levels of 12 DGs in resistant cell lines.

3.5 Bioinformatical analyses of DGs and western blot validation of some pathways

GO enrichment analysis indicated that DGs were enriched in several receptor/transporter activity-related molecular functions, including calcitonin gene-related peptide receptors, kainate selective glutamate receptors, and water transmembrane transporters (Figure A1). Some biological processes were also included in the top 30 GO terms, such as AV node cell-to-bundle of His cell signaling and SA node cell-to-atrial cardiac muscle cell communication (Figure A1). These results show that DGs are involved in the regulation of the activity of some receptors/transporters or cell signaling transduction, which may be related to chemical tolerance. KEGG classification results shown in Figure 5 indicate that >100 DGs were classified into pathways of cancer: overview, signal transduction, and immune system. Approximately 50 DGs were found to be directly involved in various cellular processes (Figure 5). Enrichment analysis indicated that the DGs were enriched in the metabolism of some molecules, including retinol, phenylalanine, and ketone bodies (Figure A2).

Figure 5

KEGG classification analysis of distribution of DGs in biological pathways.

Resistant cells showed decreased proliferation and increased migration potency. Several relevant pathways were extracted from the KEGG classification results and are listed in Table 3. In the antineoplastic pathway of drug resistance, two sub-pathways were included: platinum drug resistance and EGFR tyrosine kinase inhibitor (TKI) resistance. Twelve DGs are involved in PD-L1 expression and the PD-1 checkpoint pathway in cancer, including CD247, also known as PD-L1. Notably, 27 DGs were classified into the transcriptional misregulation pathway in cancer. Some of them encode histones (H3C3, H3-4, H3C1, and H3C7) and some are inflammatory molecules (IL6 and CXCL8). These DGs are able to regulate transcription activity in various ways, which may be closely associated with the phenotypic alteration of the resistant cells. Some growth-related sub-pathways were also included, including the p53 pathway, the cell cycle, and oocyte meiosis. CDKN2A and CDKN1A are important regulators of cell cycle progression, and may be involved in slower proliferation. Three signaling pathways were randomly selected, and the MAPK pathway had the highest number of DGs (Table 3). To verify the relationship between the pathways and resistant cells, two pathways (MAPK and PD-L1) were selected for western blot analysis. The results in Figure 6 show that the expression of PD-L1 was significantly increased in resistant cells. The expression of p-ERK1/2 in SGC996/OXA was also significantly enhanced, whereas its level did not change in GBC-SD/OXA cells. Interestingly, the expression level of ERK1/2 in GBC-SD/OXA was significantly increased; therefore, the actual level of p-ERK1/2, but not the relative level, was significantly higher than that of the sensitive cells. The increased level of PD-L1 and activated ERK pathway may be potential mechanisms for acquired resistance.

Figure 6

Western blot detection of the expression of PD-L1 and ERK1/2 pathways.