Arginine deiminase augments the chemosensitivity of argininosuccinate synthetase-deficient pancreatic cancer cells to gemcitabine via inhibition of NF-κB signaling

Diamidino-2-phenylindole (DAPI), crystal violet, Dimethyl Sulfoxide (DMSO), methyl thiazolyl tetrazolium (MTT), propidium iodide (PI), and RNase were obtained from Sigma Chemical (St. Louis, MO, USA). Bicinchoninic acid (BCA) protein assay reagent was from Pierce Chemical (Rockford, IL, USA). The ADI gene was cloned from the M. arginini genomic DNA, and the 46 kDa ADI recombinant protein (Additional file 1: Figure S1) was produced as previously described [31]. ADI activity was determined by measuring the formation of L-citrulline from L-arginine following a modified method using diacetyl monoxime thiosemicarbazide [32]. One unit of ADI activity is defined as the amount of enzyme catalyzing 1 μmol of L-arginine to 1 μmol of L-citrulline per min under the assay conditions. Finally, the measured activity of the ADI was 30 U per mg protein. GEM was purchased from Eli Lilly France SA (Fergersheim, France).

Human primary pancreatic cancer cell lines MIA PaCa-2, PANC-1, and BxPC-3, and spleen metastatic pancreatic cancer cell line SW1990, breast cancer cell lines MDA-MB-453, BT474, MDA-MB-231, and MCF-7, and hepatocellular carcinoma (HCC) cell lines HepG2 and MHCC97-H were all purchased from the American Type Culture Collection (ATCC). All cell lines were maintained in the recommended medium (HyClone, Logan, USA) containing 10% heat-inactivated fetal bovine serum (HyClone) and 1% penicillin/streptomycin (HyClone) in a humidified (37°C, 5% CO₂) incubator. Plastic wares for cell culture were obtained from BD Bioscience (Franklin Lakes, NJ).

Thirty-seven paraffin-embedded pancreatic cancer tissues were obtained from the First Affiliated Hospital of Medical College, Xi’an Jiaotong University, between 2007 and 2010. The paraffin-embedded tissue samples were then sliced into consecutive 4-μm-thick sections and prepared for immunohistochemical (IHC) studies. IHC staining was performed using an ultrasensitive SP-IHC kit (Beijing Zhongshan Biotechnology, Beijing, China), according to the manufacturer’s protocol. Briefly, after dewaxing and rehydration, the antigen was heat-retrieved, endogenous peroxidase was quenched, and the sample was blocked with 10% BSA for 30 min at room temperature. The slides were then immersed in either primary anti-ASS1 (H231; Santa Cruz Biotechnology, Santa Cruz, CA, USA) or anti-survivin (N111; Bioworld, Minneapolis, USA) rabbit polyclonal antibodies overnight at 4°C in a humid chamber, followed by rinsing and incubating with the goat anti-rabbit secondary antibody kit. The slides were stained with the 3,3-diaminobenzidine tetrahydrochloride (DAB) kit (Beijing Zhongshan Biotechnology, Beijing, China) and were subsequently counterstained with hematoxylin. Two pathologists assessed the IHC results as described previously [33]. Finally, the images were examined under a light microscope (Olympus, Tokyo, Japan). The Ethical Review Board Committee of the First Affiliated Hospital of Medical College, Xi’an Jiaotong University, China, approved the experimental protocols and informed consent was obtained from each patient who contributed tissue samples.

Total RNA from cells was prepared using trizol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol [34]. Subsequently, the total RNA was reverse-transcribed into cDNA using a Takara Reverse Transcription Kit (Takara, Dalian, China) according to the manufacturer’s recommendations. Reverse transcription-polymerase chain reaction (RT-PCR) was performed as previously described [35]. For quantitative-real time (qRT)-PCR reactions, 2 μL of cDNA was mixed with a reaction mix containing 10 μL of SYBR Green (Takara), 0.8 μL of primers, and water for a total reaction volume of 20 μL. For detecting of ASS1, caspase-3, caspase-9, Bax, Bcl-2, and survivin at mRNA levels, the following gene specific primers (Beijing Dingguo Changsheng Biotechnology) were designed as follows:

ASS1-sense: 5'-AGTTCAAAAAAGGGGTCCCT-3',

ASS1-antisense: 5'-TTCTCCACGATGTCAATACG-3';

Caspase-3-sense: 5'-GTAGAAGAGTTTCGTGAGTGC-3',

Caspase-3-antisense: 5'-TGTCCAGGGATATTCCAGAG-3';

Caspase-9-sense: 5'-GCCATGGACGAAGCGGATCGGCGG-3',

Caspase-9-antisense: 5'-GGCCTGGATGAAGAAGAGCTTGGG-3';

Survivin-sense: 5'-TCCACTGCCCCACTGAGAAC-3',

Survivin-antisense: 5'-TGGCTCCCAGCCTTCCA-3';

Bax-sense: 5'-GGCTGGACATTGGACTTC-3',

Bax-antisense: 5'-AAGATGGTCACGGTCTGC-3';

Bcl-2-sense: 5'-GTGTGGAGAGCGTCAACC-3',

Bcl-2-antisense: 5'-CTTCAGAGACAGCCAGGAG-3';

GAPDH-sense: 5'-CTCTGATTTGGTCGTATTGGG-3',

GAPDH-antisense: 5'-TGGAAGATGGTGATGGGATT-3';

The number of specific transcripts detected was normalized to the level of GAPDH. Relative quantification of gene expression (relative amount of target RNA) was determined using the equation 2^{(−ΔΔ Ct)}.

Cells were grown on glass coverslips, fixed with 4% paraformaldehyde for 10 min at room temperature, and then incubated with or without (control) the primary anti-ASS (H231) antibody overnight; the coverslips were then washed and incubated with the appropriate secondary antibody conjugated with FITC for 1 h at room temperature. DAPI was used to stain the nuclei. The coverslips were mounted onto slides, and the cells were viewed for evaluating ASS expression using a Leica TCS-SP2 confocal scanning microscope (Leica, Heidelberg, Germany).

Total protein from pancreatic cancer tissues or cells was extracted following lysis in the RIPA lysis buffer (150 mM NaCl, 50 mM Tris, 1% NP-40, 0.25% sodium deoxycholate, and 1 mM EGTA) supplemented with the protease inhibitor cocktail (Sigma, St, Louis, USA) for 30 min [36]. The resulting debris was removed by centrifugation, and the supernatant containing the protein lysate was collected. For preparation of the nuclear extracts, cells in the control and experimental groups were treated for the indicated times, then incubated on ice for 30 min followed by preparation of the nuclear extracts using a nuclear extract kit (Pierce) according to the manufacturer's instructions. The cellular protein content was determined using the BCA kit (Beyotime Biotechnology, Nantong, China), and the cell lysates were separated on a 10% SDS-PAGE gel followed by electro-transfer onto a Millipore PVDF membrane (Billerica, USA). After being blocked with 5% non-fat milk in TBST, the membranes were incubated with the respective primary antibodies (pSTAT3 [Tyr705], total-STAT3, p-ERK1/2 [Thr202/Tyr204], and ERK1/2 [all from Cell Signaling Technology, Beverly, USA]; p-Akt [Thr308], total-Akt, total NF-κB p65, caspase-3, caspase-9, XIAP, c-Jun, p21, p53, β-actin [all from Santa Cruz Biotechnology]; survivin, p-c-Jun [S73], p-NF-κB p65 [S536], lamin B1 [all from Bioworld]; cyclin D1 [Boster, Wuhan, China], and ASS1 [Proteintech, Chicago, USA]) at 4 °C overnight, followed by 1:2000 horseradish peroxidase (HRP)-conjugated secondary antibodies (anti-mouse, anti-rabbit, anti-goat; Santa Cruz Biotechnology) for 2 h. Immunoreactive bands were visualized using an enhanced chemiluminescence kit (Millipore) and photographed by GeneBox analyzer (SynGene, UK). All analyses were performed in duplicate.

Cell proliferation was determined by the MTT uptake method. Following an overnight culture in a 96-well plate in 200 μL of suitable medium, cells (5 × 10³/well) were treated with varying concentrations of ADI (0–10 mU/mL), GEM (0–10⁵ nM), or both agents for the indicated time. Then, MTT (5 mg/mL) was added and incubation was continued for 4 h, followed by termination of the reaction with 150 μL of DMSO per well. Absorbance values were determined at 490 nm on a Dias automatic microwell plate reader (Dynatech Laboratories, Chantilly, USA), using DMSO as the blank and cells cultured in untreated medium as the control group. The cell viability index was calculated using the formula of OD_sample/OD_control × 100%, while inhibition ratio calculated by formula of (1 – OD_sample/OD_control) × 100%. Each experiment was repeated three times.

Cells were seeded in a 6-well plate at a density of approximately 2.0 × 10² per well (2 mL) and allowed to attach for 24 h. Next day, the adherent cells were treated with ADI (0, or 1.0 mU/mL) or GEM (0 or 100 nM), or both. When cells were treated with both ADI and GEM, the cells were first treated with ADI (0, or 1.0 mU/mL) for 12 h, followed by 100 nM GEM for another 12 h. After a total of 24 h of treatment, cells were cultured in DMEM and incubated under optimal culture conditions for 14 days, fixed with methanol, and stained with 0.1% crystal violet. Visible colonies were manually counted and photographed.

Apoptosis was analyzed by three methods: 1) Flow cytometry: Apoptotic cells were analyzed using the Annexin-V-FITC/PI kit (BD, San Diego, USA) by a FACSCalibur flow cytometer (BD) according to the manufacturer's instructions. Briefly, cells (2 × 10⁵/well) were cultured in 6-well plates in the appropriate medium for 6 h prior to treatment with GEM (100 nM) and/or ADI (1 mU). Following incubation for the indicated times, cells were trypsinized and centrifuged, washed with PBS, and stained with Annexin V and PI in the dark. Samples were analyzed, and the percentage of apoptotic cells was evaluated. 2) In situ Annexin V/PI staining: Following the pretreatment as indicated in the flow cytometry, cells were washed with PBS and stained with 5 μL of anti-Annexin V-FITC and 5 μL of PI in 500 μL of binding buffer in the dark for 15 min and then examined using a fluorescence microscope. 3) Hoechst 33258/PI double staining: After treatment as indicated previously, cells were washed with PBS and stained with 0.1 mL of Hoechst 33258 (Beyotime Biotechnology, Nantong, China) and PI for 15 min. Stained cells were photographed under a fluorescence microscope.

Cells were harvested after treatment at different time points, and they were resuspended in PBS. The cells were then fixed in 2 mL of 70% ethanol and incubated on ice for 30 min, before being washed and treated with RNase A (100 μg/mL, 5 min) and stained with PI (50 μg/mL, 15 min). Cellular DNA content was analyzed in a Coulter Epics XL flow cytometer (Beckman-Coulter, Villepinte, France).

After the drug treatment, the cells were incubated with the NF-κB p65 (C-20, Santa Cruz Biotechnology) antibody overnight. The subsequent processing was similar to the immunofluorescence assay. At the final step, the nuclear translocation of NF-κB p65 was viewed using a confocal scanning microscope (Leica).

Six- to eight-week-old male BALB/c athymic mice were kept under pathogen-free conditions according to institutional guidelines. Each aliquot of approximately 1.0 × 10⁷ PANC-1 pancreatic cancer cells suspended in 100 μL of PBS containing 20% of Growth Factor Reduced Matrigel (Becton Dickinson Labware, Flanklin, NJ, USA) was implanted subcutaneously into the mouse flank to establish xenograft tumors. After two weeks, the mice were randomly grouped into 4 groups with six animals in each group. Mice were intraperitoneally administered either PBS (vehicle), ADI (2 U/mouse), or GEM (100 mg/kg) alone or a combination of both ADI and GEM in 100 μL of PBS every four days. The tumor size was measured every three days and the tumor volume (in mm³) was calculated using the formula V = 0.4 × D × d² (V, volume; D, longitudinal diameter; d, latitudinal diameter). Four weeks later, the mice were sacrificed and the tumors were excised and weighed. All animal experiments were conducted according to a protocol approved by the Institutional Animal Care and Use Committee of Xi’an Jiaotong University.

Statistical analyses were performed using the SPSS software (version 16.0, SPSS Inc. Chicago, USA). Experimental data in vitro and in vivo were expressed as mean ± standard deviation (SD), and were analyzed by the Student’s unpaired t-test or one-way ANOVA. For frequency distributions, a χ² test was used with modification by the Fisher’s exact test to account for frequency values less than 5. P < 0.05 was considered statistically significant.

The mRNA expression levels of ASS, a key factor that determines sensitivity to arginine deprivation via ADI, were measured in several cancer cell lines, including pancreatic cancer, breast cancer (low ASS-deficient tumor [37]), and HCC (high ASS-deficient tumor [37, 38]) using qRT-PCR. The MCF-7 breast cancer cell line was used as a standard control for identifying ASS mRNA expression. The BxPC-3 (primary pancreatic cancer), SW1990 (spleen metastatic pancreatic cancer), BT474 (breast cancer), and HepG2 (HCC) cells expressed high levels of ASS mRNA and the PANC-1 (primary pancreatic cancer), MIA PaCa-2 (primary pancreatic cancer), and MDA-MB-231 (breast cancer) cells expressed low levels of ASS relative to MCF-7 cells (Figure 1A). The MDA-MB-453 (breast cancer) and MHCC97-H (HCC) cell lines expressed similar ASS mRNA levels as MCF-7 cells. Next, the expression level of ASS protein was evaluated by western blot assay in the four pancreatic cancer cell lines. The ASS protein expression pattern was similar to the ASS mRNA expression pattern in these cells (Figure 1C). Immunofluorescence analysis verified that ASS protein was located in the cytoplasm of BxPC-3 and SW1990 pancreatic cancer cells, and similar to the qRT-PCR and western blotting results, PANC-1 and MIA PaCa-2 cells did not express substantial ASS protein in situ (Figure 1B). Furthermore, we analyzed the levels of ASS protein in 14 fresh-frozen pancreatic cancer tissue samples by western blotting and found that pancreatic cancers expressed low levels of the ASS protein (7 with ASS expression deficiency) (Figure 1D). Nine of fourteen tissue specimens were extracted for detection of ASS mRNA level, and the results show that transcriptional levels of the ASS gene were similar to its protein expression in the examined specimens (Figure 1E). Additionally, the expression of p65 (a subunit of heterodimeric NF-κB complexes) and caspase-3 (a proapoptotic protein) was evaluated in 14 pancreatic cancer tissue samples by western blotting, presenting that there was a constitutional expression of p65 and caspase-3 proteins in examined specimens, and high level of caspase-3 expression was associated with low p65 expression (r = −0.634, P = 0.027; Figure 1F).

To understand the clinical importance of ASS expression in primary human pancreatic cancer tissues, we evaluated ASS expression in human pancreatic cancer tissues by IHC method. ASS expression was detected in 19 of the 34 (56%) specimens and results from 2 of those tissues are shown (Figure 1G, i-iv). Reduced ASS expression correlated with lymph node metastasis, and local invasion in patients with pancreatic cancer (Table 1). In addition, the expression of survivin, a member of the inhibitor of apoptosis protein (IAP) family, was also detected in the same cancer specimens, and its expression was found in cytoplasm and/or nucleus in most of the cancer specimens (25/34, 74%) (Figure 1G, v and vi). By comparing the expression of ASS and survivin in pancreatic cancer specimens, a positive correlation between reduced survivin and ASS expression was exhibited (Table 2).

Next, we examined the cytotoxic effect of ADI on BxPC-3, SW1990, MIA PaCa-2, and PANC-1 cells by the MTT assay. Following treatment for one to three days, ADI significantly decreased the viability of ASS-deficient PANC-1 and MIA PaCa-2 cells in a dose- and time- dependent manner, but did not inhibit the proliferation of ASS-expressing BxPC-3 and SW1990 (Figure 2A). The 50% inhibitory concentration (IC₅₀) of ADI in PANC-1 was determined to be 1 mU/mL at 72 h and was used as the treatment dose in ensuing experiments. Subsequently, primary pancreatic cancer cell lines PANC-1 and BxPC-3 were used in further cellular and molecular experiments. The two cell lines treated with ADI or PBS were analyzed for cell cycle progression and apoptosis using FACS analysis. The findings showed that 1 mU/mL of ADI induced PANC-1 cell cycle arrest at the G1 phase and with a shorter G2/M phase, but caused scarcely any delay at the respective cell cycle phase for the BxPC-3 cell line at 24 h (Figure 2B). Similarly, 1 mU/mL ADI induced significant programmed cell death in ASS-deficient PANC-1 cells at 24, 48, and 72 h following treatment (Figure 2C), but did not cause apoptosis in ASS-positive BxPC-3 cells at 48 h (Figure 2D). In addition, the cellular morphology of PANC-1 cells was altered (Figure 2E) and the colony formation ability was attenuated upon treatment with 1 mU/mL of ADI (Figure 2F), but these cellular changes were not observed in BxPC-3 cells.

To explore the precise mechanisms of ADI-induced apoptosis in pancreatic cancer cells, we studied several apoptosis-related proteins using western blotting. Our findings showed that, after 12 h treatment, ADI significantly downregulated the expression of two IAP-family antiapoptotic proteins, namely X-linked IAP (XIAP) and survivin (Figure 3A), and simultaneously upregulated the expression of caspase-3 and caspase-9 that are responsible for the release of mitochondrial proapoptotic proteins (Figure 3B) in PANC-1 cells; however, the same concentration of ADI treatment did not alter the expression of these apoptosis-related proteins in BxPC-3 cells (Figure 3D). Next, we found considerable accumulation of p53 protein in the p53-mutant PANC-1 (Figure 3C) and BxPC-3 (Figure 3E) cells, but p53 expression was not significantly altered after ADI treatment in either cell line, and no significant change in p21 protein (a p53 induced product) expression was detected. Due to no significant cellular and molecular changes in ASS-positive BxPC-3 pancreatic cancer cell after ADI treatment, we focused on the relevant studies in the ASS-negative PANC-1 cell line. After 0 to 24 h treatment with ADI, caspase-3 activation increased progressively in a time-dependent fashion, while the expression of cyclin D1 was reduced in PANC-1 cells (Figure 4A). Furthermore, we tested the phosphorylation levels of p65 at serine 536 (p-p65 [Ser536]), shown to play a critical role in the activation of the NF-κB pathway [39, 40], in PANC-1 cells treated with ADI at several time points. We discovered that p-p65 (Ser536) decreased in a time-dependent manner following ADI treatment (Figure 4B). To understand whether ADI treatment blocked the phosphorylation of NF-κB p65 in PANC-1 cells via altering survival signaling, we detected the levels of Akt, p-Akt, ERK1/2, p-ERK1/2, STAT3, and p-STAT3. The results showed that ADI treatment for 8 h inhibited phosphorylation of STAT3 and Akt, but not ERK1/2 (Figure 4C).

To evaluate the antiproliferative activity of ADI that potentiated GEM treatment, the MTT assay was initially conducted. The IC₅₀ of GEM that inhibited the proliferation of BxPC-3 (Additional file 2: Figure S2A) and PANC-1 (Additional file 2: Figure S2B) cells was estimated by the MTT assay to be approximately 30 nM and 100 nM at 72 h, respectively; this concentration was then used in the subsequent experiments. GEM in combined with 6 h ADI pretreatment significantly inhibited the proliferation of PANC-1 cells compared to ADI or GEM alone (Additional file 2: Figure S2D), but this effect was not observed in BxPC-3 cells (Additional file 2: Figure S2C). Furthermore, by using in situ fluorescence microscopy visualization (Figure 5A) and FACS (Figure 5B), it was revealed that ADI pretreatment for 6 h promoted GEM-induced PANC-1 cell apoptosis by 24 h. We found that the apoptotic cells in situ readily stained with Annexin V-FITC/PI (green and red fluorescence) as well as with Hoechst and PI (blue and red fluorescence) (Figure 5A). Additionally, the GEM mediated S phase-arrest was enhanced in the case of pretreatment with ADI for 6 h (Figure 5C). We conducted a colony formation assay to test the colony-formation potential of individual PANC-1 cells; the results showed that GEM in combination with ADI pretreatment for 6 h reduced the colony numbers of PANC-1 cells compared to GEM or ADI alone (Figure 5D).

To further understand the molecular changes associated with ADI-mediated potentiation of GEM-induced apoptosis in PANC-1 cells, we examined the mRNA levels of Bax, Bcl-2, caspase-3, and -9, and survivin in the cells by qRT-PCR. As shown in Figure 6A, both ADI and GEM upregulated Bax, caspase-3 and -9 mRNA levels; GEM induced the mRNA level of the antiapoptotic gene Bcl-2; while ADI not only decreased the expression of Bcl-2 but also inhibited the induction of Bcl-2 transcription by GEM (with ADI pretreatment for 6 h). Additionally, ADI downregulated survivin mRNA, while GEM had limited impact on survivin gene transcription; however, the combination of ADI pretreatment for 6 h with GEM resulted in even greater inhibition of survivin expression than ADI alone.

We next determined whether ADI potentiated the GEM-induced apoptosis of PANC-1 cells by blocking activation of the NF-κB pathway. We found that ADI pretreatment for 6 h downregulated the nuclear expression of p65 and inhibited p65 induction by GEM (Figure 6B). Next, we studied whether ADI inhibited the GEM-induced p65 nuclear translocation in PANC-1 cells using in situ immunofluorescence microscopy. As shown in Figure 6C, p65 was located in the cytoplasm in the control group and in the ADI-treated group, while a significant amount of p65 was visible in the nucleus as green fluorescent spots in the GEM-treated group. However, when PANC-1 cells treated with both GEM and the pretreatment ADI for 6 h, the green fluorescent nuclear spots of p65 disappeared, indicating that ADI can block the nuclear translocation of the NF-κB p65 subunit. To evaluate whether ADI regulates GEM-induced activation of NF-κB pathway by inhibiting p65 phosphorylation, we examined the ratio of p-p65 (Ser536) to total p65 in nuclear and cytoplasmic extracts. Our analysis of the nuclear proteins showed that ADI significantly decreased p-p65 expression levels, while GEM did not. Additionally, ADI pretreatment reduced p-p65 levels in combination with GEM. In the cytoplasmic extracts, GEM significantly increased p-p65 expression which was unaffected by ADI alone, but significantly reduced in ADI pretreatment combined with GEM (Figure 6D). Together, these data provide evidence that ADI can suppress NF-κB pathway activation. Furthermore, GEM alone induced c-Jun phosphorylation at Ser73, but ADI alone or together with GEM did not. However, ADI pretreatment for 6 h could decrease the p-c-Jun induction by GEM alone, indicating that ADI could maintain c-Jun phosphorylation at the baseline level (Figure 6E). Finally, we validated the expression of survivin protein in nuclear extracts and found results similar to that for mRNA expression inhibited by ADI treatment; however, GEM increased the expression of nuclear survivin (Figure 6F).

To understanding whether ADI down-regulated the phosphorylation of NF-κB p65 by blocking activation of the PI3K/Akt survival signal pathway, we evaluated the expression of several important proteins in this signaling pathway in pancreatic cancer cells treated with ADI in combination with the PI3K inhibitor LY294002 treatment for 20 min. The results showed that ADI combined with LY294002 at 20 μM significantly down-regulated the level of p-Akt (Thr308) and p-p65 (Ser536), but not p-ERK1/2 (Thr202/Tyr204) in ASS-deficient PANC-1 cells, as compared to ADI treatment alone (P < 0.05) (Figure 7A); however, the combined treatment of ADI and LY294002 did not change the expression of relevant proteins of the PI3K/Akt and NF-κB p65 signaling pathways as compared with ADI treatment alone, in ASS-positive BxPC-3 pancreatic cancer cells (Figure 7B).

Based on our findings that ADI blocks NF-κB signaling, leading to enhance GEM-induced apoptosis of PANC-1 cells in vitro, we thus sought that it would be interesting to determine if ADI enhanced the antitumor effect of GEM in vivo. For these studies, PANC-1 cells were subcutaneouly implanted into nude mice to generate a xenograft. When the xenograft tumors had grown to approximately 50 mm³, the mice (n = 6) were treated with PBS (vehicle), ADI (2 U/mouse), GEM (100 mg/kg), or both ADI and GEM. The tumors regressed during the treatment period in all groups except the vehicle group (Figure 8A and B). From days 15–24, treatment with both ADI and GEM markedly regressed tumor growth when compared to GEM or ADI alone, and minimal tumors were observed in the combination group at resection (Figure 8B).