This article is mentioned in:

Introduction

Pancreatic cancer is a common digestive malignant tumor with low resection rate, high mortality rate and poor prognosis, as the characteristics of this tumor are masked. Pancreatic cancer patients who cannot undergo surgery are subjected to chemotherapy as a fundamental treatment modality; this modality is also a key component of systemic therapy (1). In pancreatic cancer chemotherapy, gemcitabine (GEM) was initially recommended as a first-line drug by the Food and Drug Administration (USA) in 1997. Since then, research on combination chemotherapies, such as cytotoxic drugs [5-fluorouracil (2), cisplatin (3) and capecitabine (4)] and biological agents [erlotinib (5), cetuximab (6) and bevacizumab (7)], as second-line modes of chemotherapy has been extensively conducted. Although GEM is currently the preferred drug for single chemotherapeutic applications in pancreatic cancer, the inherent and acquired resistance of cancer cells to GEM prevents the efficient improvement of the clinical benefit and survival of patients. Furthermore, the efficiency of this treatment is very low (12%) (8); as such, this drawback should be resolved in clinical applications. However, related studies have shown that the prognosis of pancreatic cancer in the past 10 years has remained unchanged.

The resistance to GEM is induced by several factors. Although numerous mechanisms have been presented, the main mechanism remains unclear. This resistance is affected by several key molecular factors, including deficiencies in drug uptake, activation of DNA repair pathways, resistance to apoptosis, enhancement of tumor microenvironments, overexpression of signaling proteins, mutations in kinase domains, activation of alternative pathways, mutations of genes and conversion to an epithelial-mesenchymal transition-like phenotype. Hence, GEM-resistance mechanisms involved in pancreatic cancer should be investigated; furthermore, a highly efficient multi-target drug with low toxicity should be developed to synergize current chemotherapy drugs or reverse drug resistance for pancreatic cancer treatment. The present study was conducted to establish a human pancreatic cancer GEM-resistant cell line and determine its biological characteristics for future studies.

Materials and methods

Cell culture and animal feeding

Human pancreatic cancer cell line PANC-1 was purchased from the Shanghai Institute of Cell Biology, China. These cells were incubated with RPMI-1640 + 10% fetal bovine serum at 37°C in a cell incubator with 5% CO2 and then digested with 0.25% trypsinogen + 2% ethylene diaminetetra acetic acid for passage at a ratio of 1:2–4 once at an interval of 2–3 days. Male nude mice were obtained from the Animal Center of the Peking Union Medical College, China. The mice were fed in a specific pathogen free-grade animal room at the Fujian Medical University Animal Center in strict accordance with aseptic principles. GEM (Hengda Pharmaceutical Co., Ltd., Shanxi, China) was dissolved in normal saline to obtain a final concentration of 100 mmol/l and stored at −20°C.

Establishment of human pancreatic cancer GEM-resistant cell line

To develop GEM-resistant PANC-1 cell line, we exposed the cells to increasing concentrations of GEM (from 50 nmol/l to 2 μmol/l) with repeated subcultures until the cells became fully resistant to GEM. Subsequently, the cells in the logarithmic phase (1/well × 50 μl) were seeded in 96-well culture plates containing 50 μl of supernatant liquid. This liquid had been used to incubate fresh mouse spleen cells for 4 days and then incubated the pancreatic cancer cells for 2 weeks to prepare the cloning culture. Single cell colonies were selected by GEM. After the cultures were cloned thrice, a stable cell clone termed PANC-1RG7 with a uniformly resistant mechanism was obtained.

Morphology and ultrastructure

Cell size and the contours of PANC-1 and PANC-1RG7 cells were observed under an optical microscope. To observe ultramicrostructure characteristics, we harvested 2×106 cells and washed them thrice with phosphate-buffered saline (PBS). Subsequently, the cells were fixed in ice-cold 4% glutaraldehyde for 2 h. The samples were subsequently fixed in 1% osmic acid for 2 h, gradually dehydrated with acetone and embedded in epoxy resin. The cells were then observed under a transmission electron microscope.

Cell growth curve

PANC-1 and PANC-1RG7 cells in the logarithmic phase (5,000/well × 1 ml) were seeded in 24-well culture plates. After 24 h of attachment, the cells were harvested and counted under an inverted microscope with 0.2% trypan blue dye. Three-wells of each cell line were monitored daily for 12 days. The cell growth curves of the 2 cell lines were drawn, and doubling time was calculated using the following equation: Td = tx24 h × [lg2/(lgNt-lgNo)], where No is the number of cells when the logarithmic growth phase began, Nt is the number of cells before cell death occurred, and t is the time between the 2 phases. Each experiment was repeated thrice.

Cell cycle analysis by flow cytometry

PANC-1 and PANC-1RG7 cells (1×106) were harvested, washed thrice with PBS and fixed in ice-cold 75% alcohol for >12 h. After fixation was completed, samples were stained with 0.005% propidium iodide for 30 min in the dark at room temperature, and then analyzed to determine the DNA content by FACSCalibur (Becton-Dickinson, Mountain View, CA, USA). Each experiment was repeated thrice.

Sulforhodamine B (SRB) assays

PANC-1 (1,000/well × 100 μl) and PANC-1RG7 (1,500/well × 100 μl) cells in the logarithmic phase were seeded in 96-well culture plates and incubated for 24 h until adherence occurred. Then, cells were treated with different concentrations of GEM, adriamycin (ADM), mitomycin C (MMC), paclitaxel (PTX), methotrexate (MTX), vincristine (VCR), gefitinib (GEF), cisplatin (DDP) and 5-fluorouracil (5-FU). Control cells were supplemented with 100 μl of RPMI-1640 culture medium. The treated cells were incubated with drugs for 96 h before SRB assays described previously (9). The dose-effect curve was plotted to calculate 50% inhibitory concentration (IC50) and resistance index (RI). Each experiment was repeated thrice.

Establishment of animal models and drug intervention

After permission of Fujian Medical University Laboratory Animal Welfare & Ethics Committee, PANC-1 and PANC-1RG7 cells (5×106 cells suspended in 200 μl of RPMI-1640) were injected percutaneously using a 29-gauge syringe with a hypodermic needle on the right shoulder back of the mice. A total of 40 integrated mice (20 in each cell line) with tumors grown to a final size of ~0.4 cm in diameter were divided into 4 groups: 10 mice with PANC-1 and 10 mice with PANC-1RG7 were included in the negative control group (injected intraperitoneally with 10 ml/kg normal saline at 15, 18, 21, 24, 27 and 30 days); 10 mice with PANC-1 and 10 mice with PANC-1RG7 were included in the GEM intervention group (injected intraperitoneally with 50 mg/kg GEM at 15, 18, 21, 24, 27 and 30 days). We observed the general conditions of the mice and tumors after they were sacrificed at 33 days.

Quantitative real-time polymerase chain reaction analysis (qPCR)

PANC-1 and PANC-1RG7 cells (3×106) were harvested. Total RNA was extracted and subjected to first-strand complementary DNA as previously described (9). qPCR was performed using the ABI prism 7500 HT sequence detection system (Applied Biosystems, Foster City, CA, USA) to detect the mRNA expression of deoxycytidine kinase (dCK), 5′-nucleotidase (NT5), cytidine deaminase (CDA), equilibrative nucleoside transporter 1 (ENT1), ENT2, ribonucleotide reductase 1 (RRM1), RRM2, DNA polymerase A (POLA), multidrug resistance protein 1 (MDR1), multidrug resistance-related protein (MRP) and breast cancer resistance protein (BCRP). The forward and reverse primers we designed are shown in Table I. Relative expression was calculated using the ΔΔCt method and our result passed the validation experiment. The results of control and treated cells are expressed as an average of the triplicate samples of at least 3 independent experiments.

Table I Sequences of polymerase chain reaction primers and sequence-specific probes of target genes and β-actin.

Table I

Sequences of polymerase chain reaction primers and sequence-specific probes of target genes and β-actin.

Gene	Forward primer (5′-3′)	Reverse primer (5′-3′)	Product length (bp)
dCK	ATCCAGCTTCCTTCTGTCATTCC	CAACGAAGTGAGAGGCACCAG	80
NT5	TAATGGTATAAACACAGGATACCATCCT	CATTATCTACTACAGCTTGCTACCTGACT	85
CDA	GAAGCGTCCTGCCTGCA	CTGGACCGTCATGACAATATACG	382
ENT1	TCTTCATGGCTGCCTTTGC	GGCTTCACTTTCTTGGGCC	79
ENT2	CAAGACCTCATGGAAAGGGTG	CCACTCTGAACCCTCTGGTCA	124
RRM1	GGCACCCCGTATATGCTCTA	CCAGGGAAGCCAAATTACAA	148
RRM2	GGCTCAAGAAACGAGGACTG	TCAGGCAAGCAAAATCACAG	93
POLA	GGCTCGGATCTGTGAACCAA	GGGCTCCATATCTGTTCCCG	256
MDR1	AGGTTCCAGGATTGGCGTCTT	CCAGTCATTGCTGCGGTTTCA	156
MRP	GCGAGTGTCTCCCTCAAACG	TCCTCACGGTGATGCTGTTC	118
BCRP	GATATGGATTTACGGCTTTGC	CGATGCCCTGCTTTACCAA	135
β-actin	AGTGTGACGTGGACATCCGCAAAG	ATCCACATCTGCTGGAAGGTGGAC	220

[i] dCK, deoxycytidine kinase; NT5, 5′-nucleotidase; CDA, cytidine deaminase; ENT1, equilibrative nucleoside transporter 1; RRM1, ribonucleotide reductase 1; POLA, DNA polymerase A; MDR1, multidrug resistance protein 1; MRP, multidrug resistance-related protein; BCRP, breast cancer resistance protein.

Western blotting

PANC-1 and PANC-1RG7 (9×106) cells were harvested. Total protein fractions were extracted, separated on SDS-PAGE and then exposed to specific antibodies using western blotting described earlier (9). The specific primary antibodies we used were mouse monoclonal antibodies anti-human β-actin (sc-47778), ENT1 (sc-377283), ENT2 (sc-373871), NT5 (sc-32299), POLA (sc-137021), p-gp (sc-55510) (all from Santa Cruz, USA), and MRP (no. ab32574; Abcam, USA); rabbit polyclonal antibodies anti-human DCK (no. ab151966; Abcam, USA), CDA (sc-134754), BCRP (sc-130933) (both from Santa Cruz), Akt (no. BS1810) and mTOR (no. BS3611) (both from BioWorld, USA); rabbit monoclonal antibodies anti-human PI3K (no. 4249; Cell Signaling Technology, USA); and goat polyclonal antibodies anti-human RRM1 (sc-11733) and RRM2 (sc-10846) (both from Santa Cruz). Images were analyzed using Quantity One 4.62. Each experiment was repeated >3 times.

Statistical analysis

Experimental data are presented as the means ± standard deviation (SD) and analyzed by SPSS 19.0. Comparisons were performed using Student’s t-test between 2 groups. P<0.05 was considered to indicate a statistically significant difference.

Results

Morphological and ultrastructure characteristics

PANC-1 cells appeared fusiform under an optical microscope at a magnification of ×200. During GEM intervention, the cells appeared polygonal with elongated pseudopodia and growth retardation; the size significantly increased and numerous vacuoles were formed in the cytoplasm. Cell growth was gradually restored after GEM was removed; a fusiform was formed but remained smaller and grew more slowly than parental cells (Fig. 1A).

Figure 1 (A) Morphological changes in pancreatic cancer PANC-1 cells before, during and after GEM intervention under an optical microscope. (B) Ultrastructural changes in pancreatic cancer PANC-1 and PANC-1RG7 cells under a transmission electron microscope. GEM, gemcitabine.

PANC-1 cells contained intact a cell membrane and nucleus, numerous microvilli on the membrane, abundant organelles and a satisfactory state under a transmission electron microscope. However, PANC-1RG7 exhibited different ultrastructural characteristics. In particular, small vacuoles and lipid droplets were formed in the cytoplasm. The number of glycogen granules and lysosomes increased significantly, the mitochondrial cristae was broken as vacuolization occurred, and the rough endoplasmic reticulum became swollen (Fig. 1B).

Cell growth curve

Compared with the parental cell PANC-1, GEM-resistant pancreatic cancer PANC-1RG7 cells slowly grew at a significant rate (Fig. 2). The doubling times of PANC-1 and PANC-1RG7 cells were 25.83±2.03 and 33.83±2.15 h, respectively. The doubling time of PANC-1RG7 cells was significantly increased (p<0.05).

Figure 2 Cell growth curves of PANC-1 and PANC-1RG7 cells. PANC-1RG7 cells grew significantly more slowly than the parental PANC-1 cells.

Cell cycle analysis by flow cytometry

In PANC-1, 68.98±2.32 and 18.02±0.63% of the cells were detected in the G0/G1 and S phase, respectively. In PANC-1RG7, 69.23±3.03 and 17.77±0.89% of the cells were detected in the G0/G1 and S phase, respectively. No significant difference was determined (p>0.05; Fig. 3).

Figure 3 Flow cytometric analysis of the cell cycle. No significant difference was observed between PANC-1 and PANC-1RG7.

SRB assays

We detected 9 common chemotherapeutics, including GEM, ADM, MMC, PTX, MTX, VCR, GEF, DDP and 5-FU. The IC50 values of GEM, MTX, GEF, DDP and 5-FU were statistically different between PANC-1 and PANC-1RG7; by contrast, the IC50 values of the other drugs were not different (Table II). The RIs of GEM, MTX, GEF, DDP and 5-FU were 39.9, 2.24, 1.42, 2.35 and 7.00, respectively. This result indicated that the PANC-1RG7 cells established in this study expressed resistance to GEM and cross-resistance to MTX, GEF, DDP and 5-FU.

Table II SRB assay results of PANC-1 and PANC-1RG7 human pancreatic cancer cells treated with various concentrations of GEM, ADM, MMC, PTX, MTX, VCR, GEF, DDP and 5-FU for 96 h.

Table II

SRB assay results of PANC-1 and PANC-1RG7 human pancreatic cancer cells treated with various concentrations of GEM, ADM, MMC, PTX, MTX, VCR, GEF, DDP and 5-FU for 96 h.

IC50	PANC-1	PANC-1RG7	P-value
GEM (μmol/l)	0.0081±0.0014	0.3233±0.0933	0.0023a
ADM (nmol/l)	10.2380±2.4875	9.1433±2.4533	0.7011
MMC (μmol/l)	0.5689±0.5180	0.2545±0.1543	0.4972
PTX (nmol/l)	2.2618±0.2262	2.4840±0.1500	0.2682
MTX (nmol/l)	0.9346±0.1649	2.0948±0.1672	0.0199b
VCR (nmol/l)	7.0513±2.3578	5.4185±1.7090	0.4265
GEF (μmol/l)	7.2575±0.5216	10.281±0.2890	0.0189b
DDP (μmol/l)	0.6317±0.2159	1.4853±0.4649	0.0108b
5-FU (μmol/l)	1.9298±0.4420	13.509±2.7563	0.0001a

{ label (or @symbol) needed for fn[@id='tfn2-or-33-01-0383'] } Data shown are IC₅₀ (mean ± SD).

a Compared with PANC-1, p<0.01;

b compared with PANC-1, p<0.05.

{ label (or @symbol) needed for fn[@id='tfn5-or-33-01-0383'] } GEM, gemcitabine; ADM, adriamycin; MMC, mitomycin C; PTX, paclitaxel; MTX, methotrexate; VCR, vincristine; GEF, gefitinib; DDP, cisplatin; 5-FU; 5-fluorouracil; SRB, sulforhodamine B.

Establishment of animal models and drug intervention

We successfully established nude mouse subcutaneous tumor models. The mice were sacrificed at 33 days. Subcutaneous tumors were completely peeled off and weighed (Table III). PANC-1 tumors in the negative control group were significantly smaller than PANC-1RG7 tumors (p<0.05). This result indicated that PANC-1RG7 cells grew faster in vivo than PANC-1 cells. A significant difference was observed in tumor weights before and after GEM intervention was administered in PANC-1 (p<0.05), but not in PANC-1RG7 (p<0.05). The inhibition rates of GEM in PANC-1 and PANC-1GR7 were 82.03 and 33.40%, respectively. This finding indicated that the inhibition of GEM decreases in vivo.

Table III Subcutaneous tumor weight of each group of mice with pancreatic cancer (n=10, mean ± SD).

Table III

Subcutaneous tumor weight of each group of mice with pancreatic cancer (n=10, mean ± SD).

Tumor weight (g)	Control	GEM (50 mg/kg)
PANC-1	0.2118±0.0521	0.0381±0.0215a
PANC-1RG7	0.3247±0.1292b	0.2163±0.0833

{ label (or @symbol) needed for fn[@id='tfn6-or-33-01-0383'] } Subcutaneous tumors were completely peeled off after the mice were sacrificed at 33 days.

a Compared with control group, p<0.05;

b compared with PANC-1, p<0.05.

{ label (or @symbol) needed for fn[@id='tfn9-or-33-01-0383'] } GEM, gemcitabine.

qPCR

We examined the expression levels of dCK, NT5, CDA, ENT1, ENT2, RRM1, RRM2, POLA, MDR1, MRP and BCRP at mRNA levels by qPCR. However, only CDA, MRP and BCRP expressions changed at mRNA levels. The expression levels of these 3 genes in PANC-1RG7 were lower than those in PANC-1 with a significant difference (Table IV).

Table IV mRNA expression of dCK, NT5, CDA, ENT1, ENT2, RRM1, RRM2, POLA, MDR1, MRP and BCRP by qPCR.

Table IV

mRNA expression of dCK, NT5, CDA, ENT1, ENT2, RRM1, RRM2, POLA, MDR1, MRP and BCRP by qPCR.

PANC-1RG7 gene	ΔΔCt	2−ΔΔCt
dCK	0.06±0.52	1.00±0.33
NT5	0.62±1.33	0.86±0.77
CDA	5.88±0.69	0.02±0.01a
ENT1	0.70±0.22	0.62±0.09
ENT2	0.38±0.68	0.83±0.36
RRM1	0.04±0.43	1.00±0.29
RRM2	−0.04±0.40	1.05±0.30
POLA	0.45±1.55	0.98±0.67
MDR1	1.27±1.44	0.55±0.46
MRP	1.48±0.28	0.36±0.07a
BCRP	2.42±0.34	0.19±0.05a

{ label (or @symbol) needed for fn[@id='tfn10-or-33-01-0383'] } PANC-1 and PANC-1RG7 human pancreatic cancer cells were harvested and total RNA was extracted. Only the mRNA expression of CDA, MRP and BCRP decreased in PANC-1RG7 (2^−ΔΔCt <0.50).

a 2^−ΔΔCt <0.5.

{ label (or @symbol) needed for fn[@id='tfn12-or-33-01-0383'] } dCK, deoxycytidine kinase; NT5, 5′-nucleotidase; CDA, cytidine deaminase; ENT1, equilibrative nucleoside transporter 1; RRM1, ribonucleotide reductase 1; POLA, DNA polymerase A; MDR1, multidrug resistance protein 1; MRP, multidrug resistance-related protein; BCRP, breast cancer resistance protein.

Western blotting

We further examined the expression levels of dCK, NT5, CDA, ENT1, ENT2, RRM1, RRM2, POLA, MDR1, MRP and BCRP at protein levels by western blotting. The key components of the PI3K/Akt/mTOR signaling pathway were also determined. The examined proteins, except BCRP, were expressed in PANC-1 and PANC-1RG7. Compared with those in parental PANC-1, NT5, RRM1 and RRM2 expression levels were significantly increased in PANC-1RG7 (p<0.05). No changes in other proteins at a protein level were noted (Fig. 4).

Figure 4 Western blotting for the resistance-related factors. The examined proteins, except BCRP, were expressed in PANC-1 and PANC-1RG7. Compared with those of the parental PANC-1 cells, NT5, RRM1 and RRM2 expression levels were significantly increased in PANC-1RG7 (p<0.05). No changes in other proteins at the protein level were noted. Results represent 3 independent experiments. Gray scales shown at the bottom indicate results compared with β-actin.

Discussion

Chemotherapy is the fundamental treatment modality for pancreatic cancer patients who are unable to undergo surgery; this modality is also a key component of systemic therapy (1). GEM is currently the preferred drug for the treatment of pancreatic cancer by single chemotherapeutic applications. However, the inherent and acquired resistance of cancer cells to GEM limits its efficiency. Thus far, no effective drug has improved the clinical benefits of GEM. Hence, resistance to GEM remains a vital problem. Numerous mechanisms have been presented, however, the main one remains unclear. In the present study, a stable human pancreatic cancer GEM-resistant cell line was established for use in further studies on GEM resistance.

Intermittent intervention in gradually increasing concentration or pulse intervention in large concentrations can be performed to establish drug-resistant cancer cell lines. The former method can be used to simulate the intermittent administration of drugs in clinical applications with a high achievement ratio and increased stability. Although this method requires time-consuming procedures, we performed this method in the present study. In China, human pancreatic cancer cell line SW1990, which is derived from pancreatic cancer accompanied by metastatic spleen, is commonly used to establish a pancreatic cancer GEM-resistant cell line (10–12), while MIA PaCa-2 derived from pancreas tissues is used in other countries (13–15). However, we chose human pancreatic cancer cell line PANC-1, which is derived from pancreatic ductal carcinoma. Commonly used in clinical practice, this cell line is highly sensitive to GEM owing to its low differentiation. The stable cell clone termed PANC-1RG7 with a uniform resistant mechanism was obtained after GEM intervention was conducted for 2 years and clone cultures were prepared thrice. Our results showed that the RI of GEM was 39.9, indicating low resistance.

Changes in the morphological characteristics of resistant cells indicate acquired resistance. The established GEM-resistant cells were smaller and grew more slowly than the parental cells. Cell organs, such as lysosomes, mitochondria and rough endoplasmic reticulum, significantly changed, as observed under a transmission electron microscope. These changes may be considered the basis of functional changes related to resistance mechanisms. Multidrug resistance (16), characterized by cells that are resistant not only to intervention drug but also to other chemotherapeutics without structural or functional relationships existed in PANC-1RG7 with cross-resistance to MTX, GEF, DDP and 5-FU, suggesting that a common mechanism could be implicated in this resistance. As such, in vivo studies involving xenografts are necessary to detect biological behavior and tumor characterization. The established PANC-1RG7 cells indicated an increased discernible invasion and growth compared with parental PANC-1 cells; by contrast, in vitro studies showed a slow growth. Further studies should be conducted to investigate the possible mechanism implicated in the difference between in vivo and in vitro processes.

ENTs facilitate the entry of GEM, a pyrimidine analog, across the plasma membrane of cells. GEM is then phosphorylated intracellularly by dCK via multiple steps to derive diphosphate and triphosphate. The former is an active metabolite inhibiting RRM, resulting in a decrease in intracellular dCTP; thus, DNA synthesis is suppressed. The latter inhibits DNA synthesis by interfering with the incorporation of endogenous dCTP into DNA. Studies on GEM resistance mechanisms have shown that factors involved in GEM metabolism and transport are related to resistance. In the present study, the expression levels of the main factors in PANC-1RG7 and parental PANC-1 were detected. CDA expression at an mRNA level significantly decreased in PANC-1RG7, yet it remained unchanged at a protein level. The expressions of NT5, RRM1 and RRM2 proteins were significantly increased in PANC-1RG7 compared with those in PANC-1RG7. By contrast, no change was observed at the mRNA level. No linear relationship between mRNA and protein expression was noted since mRNA undergoes a series of regulatory processes, including microRNA regulation, translation, post-translational modification (e.g., glycosylation and phosphorylation), and protein transport, to express proteins. Thus, the differences in the changes between mRNA and protein expression could be acceptable. As active proteins, enzymes are involved in activities more directly related to expressions at protein levels. Hence, the overexpression of NT5, RRM1 and RRM2 was necessary to induce the resistance of the established PANC-1RG7 to GEM.

Increased activities of RRM1 and RRM2 possibly promote the conversion of nucleoside to deoxynucleoside and accelerate DNA polymerization and repair, resulting in resistance (17). In GEM metabolism and transport, 4 factors, ENT1, dCK, RRM1 and RRM2, are implicated in acquired resistance (18). As ENT1 expression decreases, GEM intake is reduced and cytotoxicity is decreased in vivo (19). The deficiency in dCK activities is one of the mechanisms by which pancreatic cancer cells develop resistance to chemotherapeutic drugs (20) since dCK is an important factor in the intracellular conversion of GEM to an active metabolite. However, ENT1 and dCK expression in PANC-1RG7 remained unchanged at the mRNA and protein levels. Enzyme activity is not only affected by mRNA or protein expression; studies have shown that dCK activity, protein and gene expression levels are significantly correlated (21). No dCK activity was detected directly due to limited experimental conditions. However, dCK failed to induce PANC-1RG7 to develop resistance to GEM. In a previous study, the overexpression of NT5 was observed in a human pancreatic cancer GEM-resistant cell line (14); however, further studies should be conducted to determine whether or not this overexpression increases the removal of GEM.

P-gp, MRP and BCRP are 3 multidrug-resistant proteins relevant to tumor stem cell. On the basis of the results of expression detection, we found that P-gp and MRP proteins were expressed in human pancreatic cancer PANC-1 cells; MDR1, MRP and BCRP genes were also expressed in PANC-1 cells, indicating the inherent resistance of PANC-1. The expressions did not increase after GEM intervention was administered in the present study, and the gene expression of MRP and BCRP decreased. However, studies have yet to determine whether or not the resistance of cancer cells to chemotherapeutics is attributed to the decrease in the gene expression of MRP and BCRP. Further studies are required to determine if these changes are correlated with cell resistance. Nevertheless, the proteins and genes not implicated in the resistance of PANC-1RG7 to GEM could be identified. Moreover, changes in apoptotic signaling pathway are related to the resistance of pancreatic cancer cells. Thus, apoptosis-regulatory proteins are abnormally expressed in pancreatic cancer cells (22). No changes in PI3K, Akt and mTOR protein expression were observed in PANC-1RG7.

In summary, human pancreatic cancer GEM-resistant cell line PANC-1RG7 was established in this study. These cells grew slowly in vitro but rapidly in vivo. In vitro and in vivo experimental results showed that PANC-1RG7 exhibited stable resistance to GEM and cross-resistance to other chemotherapeutics, such as MTX, GEF, DDP and 5-FU. Of all the factors related to GEM resistance, only RRM1 and RRM2 protein expression increased in the resistant cells, thereby inducing resistance to GEM. This result indicated that the overexpression of RRM1 and RRM2 was necessary to induce the resistance of PANC-1RG7 to GEM.

Cancer-resistant cell lines established in vitro remain the main tools used to study the mechanisms of acquired tumor resistance. The established pancreatic cancer GEM-resistant cell line PANC-1RG7 may also be used as an important tool to investigate the acquired resistance of pancreatic cancer, RRM, or new chemotherapy drugs that can reverse GEM resistance.

Acknowledgements

The present study was supported by the National Science Foundation of China (no. 30772587), the Great Research Project of Fujian Medical University (no. 09ZD012), and the Natural Science Foundation of Fujian Province (nos. C0510012 and 2011J01188).

References