Journal list menu
Cannabinoid derivatives induce cell death in pancreatic MIA PaCa-2 cells via a receptor-independent mechanism
Abstract
Cannabinoids (CBs) are implicated in the control of cell survival in different types of tumors, but little is known about the role of CB system in pancreatic cancer. Herein, we investigated the in vitro antitumor activity of CBs and the potential role of their receptors in human pancreatic cancer cells MIA PaCa-2. Characterization tools used for this study included growth inhibition/cell viability analyses, caspase 3/7 induction, DNA fragmentation, microarray analysis and combination index-isobologram method. Our results demonstrate that CBs produce a significant cytotoxic effect via a receptor-independent mechanism. The CB1 antagonist N-(piperidin-1-1yl)-5-(4-iodophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide (AM251) was the most active compound with an IC50 of 8.6 ± 1.3 μM after 72 h. AM251 induces apoptosis, causes transcriptional changes of genes in janus kinase/signal transducers and activators of transcription signaling network and synergistically interacts with the pyrimidine analogue, 5-fluorouracil. These findings exclude the involvement of CB receptors in the regulation of MIA PaCa-2 cell growth and put AM251 forward as a candidate for the development of novel compounds worthy to be tested in this type of neoplasia.
1 Introduction
Several findings support the role of cannabinoids (CBs) and their derivatives in cell growth inhibition and apoptosis induction in tumor cells [1, 2]. The plethora of experimental observations regarding the mechanism involved in the antiproliferative action of cannabinoids is in line with their pleiotropic nature. From a molecular viewpoint, receptor-ligand binding appears to play a key role in CB-mediated cytotoxic effects, as demonstrated for WIN-55,212-2 (mixed CB1/CB2 receptors agonist) on human prostate cells LNCaP [3]. On the contrary, several observations on cancer cells which express both CB1 and CB2 receptors underline the irrelevance of CB receptor status in determining drug response [4-6]. For instance, the ability of cannabidiol to impair the migration of U87 glioma cells was not antagonized by the selective receptor antagonists SR141716 (CB1) and SR144528 (CB2) [4], while δ9-tetrahydrocannabinol (THC)-mediated apoptosis in leukemic cell lines [5] and in human prostate PC-3 cells [6] was independent of the CB receptors. Moreover, immunohistochemical and functional analyses in mouse models of glioma [7] and skin carcinoma [8] have shown that CBs may exert their antitumor effects by blocking the angiogenic process.
Despite the above mentioned observations, we need to be cautious when envisaging the potential clinical use of new anticancer therapies. As a matter of fact, nanomolar concentrations of δ9-tetrahydrocannabinol lead to epidermal growth factor receptor-mediated increase in glioblastoma and lung carcinoma cell proliferation [9], suggesting that biological responses to CBs critically depend on drug concentration and type of cell examined.
The aim of the present study was to investigate the effect of CBs on the human pancreatic cancer cells MIA PaCa-2 since, to date, no data have been produced on the putative role of CB system in this specific cancer cell type.
2 Materials and methods
2.1 Chemicals
The following CBs were used: WIN-55,212-2, CB1/CB2 receptor agonist; ACEA, CB1 receptor-selective agonist; JWH-015, CB2 receptor-selective agonist; AM251 and SR141716A, CB1 receptor-selective antagonists; AM630, CB2 receptor-selective antagonist (Tocris, Northpoint, UK). 5-fluorouracil (5-FU; Sigma–Aldrich, St. Louis, MO, USA) was used as reference drug. The reversible inhibitor of caspase 3/7, 5-[(S)-(+)-2-(methoxymethyl)pyrrolidino]sulfonylisatin, was obtained from Calbiochem (La Jolla, CA, USA). Compounds were dissolved in dimethyl sulfoxide (DMSO) and further diluted in sterile culture medium immediately before their use.
2.2 Cell culture
The human pancreatic cancer cell line MIA PaCa-2 was obtained from the American Type Culture Collection (ATCC, Rockville, MD) and cultured in DMEM medium supplemented with l-glutamine (2 mM), 10% fetal bovine serum, 2.5% horse serum, 50 IU/ml penicillin and 50 μg/ml streptomycin (Sigma–Aldrich, Milano, Italy) at 37 °C in an atmosphere of 5% CO2.
2.3 Growth inhibition assay
Cells (8 × 103/well) were seeded in 24-well plate and incubated for 24 h at 37 °C to adhere completely to plastic and start growing. After that, the medium was changed and cells were treated with CBs at 0.01–100 μM for 24, 48 and 72 h. Cells were harvested and counted by haemocytometer, and the growth of treated cells was expressed as a percentage of the untreated control cells. The concentration of drugs that decreased cell count by 50% (IC50) was calculated by nonlinear least-squares curve fitting of experimental data (GraphPad Software, San Diego, CA, USA).
2.4 Cell viability assay
Cell viability was measured using a method based on the cleavage of the 4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate (WST-1) to formazan by mitochondrial dehydrogenase activity (Cell proliferation reagent WST-1; Roche, Mannheim, Germany). Cells (2 × 103/well) were seeded in 96-well plate in medium containing 20% FBS and 2.5% HS, 1% FBS and 0.25% HS or serum-free medium and received CBs from 0.14 to 312 nM, depending on their K i for CB receptors, and at 10 μM for 24, 48 and 72 h. Following drug exposure, WST-1 was added and the absorbance was measured at 450 nm using a microplate reader.
2.5 RT-PCR characterization of CBs receptors
RNA from Mia PaCa-2 cells was extracted by using the SV total RNA isolation system (Promega, Madison, WI, USA), reverse-transcribed by the OmniScript RT kit and PCR was performed by the HotStartTaq Master Mix kit (Qiagen, Valencia, CA, USA). cDNA samples were amplified for 35 cycles: 1 min at 95 °C; 1 min at 56.8 °C (CB1), 58 °C (CB2) or 55 °C (GAPDH); 1 min at 72 °C, plus elongation step at 72 °C for 10 min. Primers used were: 5′-CCACTCCCGCAGCCTCCG-3′(F) and 5′-ATGAGGCAAAACGCCACCAC-3′(R) for CB1; 5′-CGCCGGAAGCCCTCATAC-3′(F) and 5′-CCTCATTCGGGCCATTCTTG-3′(R) for CB2; 5′-GTGAAGGTCGGTGTCAACG-3′(F) and 5′-GGTGAAGACGCCAGTAGACTC-3′(R) for GAPDH and the expected amplification products were 293, 522 and 299 bp long, respectively. CB1 and GAPDH primers were designed by Oligo-Primer Analysis Software 4.0 (http://www.oligo.net/) whereas CB2 primers sequences were retrieved from McKallip et al. [10]. Human lymphocytes-derived genomic DNA from a healthy volunteer was used as positive control for testing CB1 and CB2 primers efficiency.
2.6 Caspase activity assay
Enzyme activity was assessed by the Apo-ONE™ Homogeneous Caspase-3/7 assay substrate (Promega, Madison, WI, USA). Cells were seeded at 2–7 × 103/wells and treated with CBs or 5-FU at 10 μM for 24, 48 and 72 h. Subsequently, the Caspase-3/7 assay substrate was added and the fluorescence was measured by spectrofluorimeter at excitation and emission wavelengths of 485 and 530 nm, respectively. Values were expressed as ratio between fluorescent signals generated in cells treated with CBs or 5-FU and those produced in untreated cells (vehicle alone).
2.7 DNA fragmentation assay
Cells were treated at AM251 10, 50 and 100 μM for 24, 48 and 72 h. Cells harvested by trypsinization were combined with detached cells and apoptosis was assessed by the Cell Death Detection ELISA kit (Roche, Mannheim, Germany) based on the recognition of released nucleosomes after DNA internucleosomal fragmentation by a mouse monoclonal antibody directed against DNA and histones.
2.8 Microarray assay
Total RNA was isolated from AM251 (10 μM for 24 h)-treated and untreated (vehicle alone) MIA PaCa-2 cells by using the RNeasy purification kit (Qiagen, Valencia, CA, USA). DNaseI treated RNA was amplified and dye-labeled with Cy3 or Cy5 by using the Amino Allyl MessageAmp™ II aRNA Amplification Kit (Ambion, Austin, TX, USA). Competitive hybridizations on high-density oligonucleotide microarray slides, containing 21 329 sequences from the Operon 2.0 oligoset (Microcribi, Padua, Italy), were automatically performed by Lucidea Automatic Slide Processor (GE Healthcare Technologies, UK). We ran three independent experiments, each time in “dye swap” to eliminate potential bias due to differences in dye labelling efficiency, for a total of six hybridizations.
2.9 Microarray data acquisition and analysis
Hybridized slides were scanned with Genepix 4000B scanner using GenepixPro 4.0 software (Molecular Devices, Sunnyvale, CA, USA). Resulting files were analyzed within the R environment by making use of the “Limma” package (http://www.bioconductor.org). The loaded data were filtered by adding additional “Bad” flags to those spots which had not an acceptable signal to noise ratio (SNR); the “Bad” spots were discarded from the successive analysis. The print-tip normalization was applied to the data and the Bayesian statistics B were calculated for each gene with the “Limma” package (http://www.bioconductor.org).
The list of differentially regulated genes was analyzed by Onto-Express and Pathway-Express tools (http://vortex.cs.wayne.edu/projects.htm) to find significantly impacted biological processes.
2.10 Real-Time PCR validation of microarray data
Total RNA was reverse transcribed by SuperScript III kit (Invitrogen, Carlsbad, CA, USA). Primer sequences were designed using Beacon Designer 4.0 software (Biorad, Hercules, CA, USA) and are available upon request. PCR amplifications were performed with Platinum Sybr Green QPCR Supermix (Invitrogen, Carlsbad, CA, USA) by using an iCycler iQ Real-Time PCR detection system (Biorad, Hercules, CA, USA). For each sample, expression of the human cytochrome c-1 (CYC1) gene was used to normalize the amount of the investigated transcript. Relative expression levels were calculated with respect to the expression level in the untreated sample and were expressed as fold increase or decrease.
2.11 Quantitation of the synergism between AM251 and 5-FU
Cells were seeded at 8 × 103/well and treated with 5-FU in the range of 0.5–75 μM alone or in combination with AM251 at 1, 10 and 50 μM for 72 h. The combination index-isobologram method [11] was applied to analyze data from the in vitro drug-combination study. Briefly, synergism or antagonism for AM251 plus 5-FU was calculated on the basis of the multiple drug-effect equation, and quantified by the combination index (CI), where CI < 1, CI = 1, and CI > 1 indicate synergism, additive effect, and antagonism, respectively. Based on the classic isobologram for mutually exclusive effects, CI values were calculated as follows: CI = [(D)1/(D x )1] + [(D)2/(D x )2], where (D)1 and (D)2 are the concentrations of AM251 and 5-FU in combination that induce a x% of cell growth inhibition (isoeffective as compared with the single drugs alone) while, (D x )1 and (D x )2 are the concentrations for x% inhibition by AM251 and 5-FU, respectively. D x values were obtained from the following equation: (D x ) = D m[f a/(1 − f a)]1/m , where D m is the median-effect dose (ED50) of the single drug, f a is the fractional inhibition, (1 − f a) is the fraction unaffected, and m is the coefficient signifying the shape of the dose–effect curves.
The dose-reduction index (DRI) defines the extent (fold change) of concentration reductions that are possible in a combination schedule for a given degree of effect as compared with the concentration of each drug alone as follows: (DRI)1 = (D x )1/(D)1 and (DRI)2 = (D x )2/(D)2. The DRI values for actual combination data points were calculated from the results obtained from CI equation.
2.12 Statistical analysis
Data were presented as mean values ± standard error (S.E.). Statistical comparisons among groups were performed by Student's t-test or one-way analysis of variance (ANOVA) followed by the Newman–Keuls test for multiple comparison. Significance was assumed at P < 0.05.
3 Results
3.1 In vitro antitumor activity of CBs and role of CB receptors
Irrespective of their pharmacological profiles on CBs receptors, CBs produced a concentration- and time-dependent MIA PaCa-2 cell growth inhibition (Table 1 ). Accordingly, CBs at 10 μM for 24–72 h led to a time-dependent drop in cell viability; such an effect was directly related to the serum concentration and AM251 was the most active compound within CBs tested (Fig. 1 A). WIN-55,212-2 and SR141716A, two other structurally-unrelated cannabinoid receptor ligands, did not affect significantly cell viability at 10 μM for 24–72 h (Fig. 1B) and therefore no further investigation was performed on these compounds.
| Compounds | IC50 (μM) | ||
|---|---|---|---|
| 24 h | 48 h | 72 h | |
| ACEA | 48.2 ± 3.4 | 23.9 ± 8.7 | 16.9 ± 1.7 |
| AM251 | 44.4 ± 3.8 | 18.7 ± 2.9 | 8.6 ± 1.3∗ |
| JWH-015 | 42.4 ± 2.4 | 20.0 ± 2.0 | 17.0 ± 1.4 |
| AM630 | 64.6 ± 1.9 | 63.5 ± 8.7 | 19.2 ± 1.7 |
Experiments were performed in triplicate and values are expressed as means ± S.E.; ∗ P < 0.05, as compared to the IC50 of the other cannabinoids after 72 h.
Fig. 1C shows the RT-PCR product corresponding to CB1 receptor, whose identity was confirmed by DNA sequencing. The mRNA of CB2 receptor was not detected on MIA PaCa-2 cells, although the positive control for CB2 receptor on human lymphocytes-derived genomic DNA clearly demonstrated the quality of CB2 primers. Furthermore, at concentrations 0.1–10-fold their K i for CB receptors, both the selective CB1 receptor agonist and antagonist, ACEA and AM251, did not significantly affect cell viability both in normal-serum and serum-free medium (Fig. 1D). Finally, the antiproliferative effect of ACEA was not prevented by AM251 at the non-cytotoxic concentration of 1 μM (Fig. 1E).
3.2 Effect of CBs on apoptosis
At variance with the other CBs as well as the reference drug, 5-FU, the magnitude of AM251 effect on caspase 3/7 was related to the time of exposure to the drug. Particularly, AM251 at 10 μM for 24 h induced enzyme activities (1.78 ± 0.22 fold stimulation vs. control; P < 0.01) which were further enhanced at 3.01 ± 0.34 and 8.18 ± 0.34 fold stimulation (P < 0.001) after 48 and 72 h, respectively (Fig. 2 A). Fig. 2B demonstrates that 5-[(S)-(+)-2-(methoxymethyl)pyrrolidino]sulfonylisatin, a reversible inhibitor of caspase 3/7, partially reverses the toxicity of AM251. Finally, AM251 at 10, 50 and 100 μM for 24–72 h, significantly increased the production of cytosolic histone-associated DNA fragments in a concentration- and time-dependent manner (Fig. 2C).
3.3 Signaling pathways involved in AM251 antitumor activity
Thirty-nine genes were downregulated while 46 genes were upregulated after AM251 at 10 μM for 24 h. Onto-Express automatically translated such a list of differentially regulated genes into functional profiles with the result of 63.5% of genes assigned to “cellular process” category which, in turn, includes genes involved in cell death, cell cycle and regulation of cell proliferation. Furthermore, pathway-express analysis indicated that AM251 treatment induced a modulation of genes of highly-regulated signaling networks critical for cell fate including janus kinase/signal transducers and activators of transcription (JAK-STAT) and mitogen-activated protein kinase (MAPK) signaling pathways and cell cycle, as well. Fig. 3 A shows a multilevel molecular targeting which may represent a conceivable hypothetical framework of the AM251 mechanism of action.
The complete list of the differentially expressed genes is available in supplementary material. Further information can be retrieved from the ArrayExpress database at the European Bioinformatics Institute (EBI) (http://www.ebi.ac.uk/arrayexpress/) by using the following accession numbers: E-MAXD-9 for the experiment and A-MAXD-4 for the array. Microarray results were validated by quantitative real-time PCR which confirmed the AM251-induced changes of a representative cohort of genes (Fig. 3B shows some examples of genes validated by real-time PCR).
3.4 Synergistic interaction between AM251 and 5-FU
Fig. 4 A shows that, in the presence of AM251 at the non-cytotoxic concentration of 1 μM (Fig. 4B), the dose–response curve of 5-FU shifted to the left with the result of a significant reduction (P < 0.01) of the IC50 value from 6.0 ± 0.2 to 3.7 ± 0.2 μM (5-FU alone and in combination schedule, respectively). Sample calculation for the combination index (CI) values of AM251 at 1 μM plus 5-FU from 0.5 to 5 μM, demonstrated synergism (CI < 1) and a favorable dose-reduction index (DRI > 1) for both the drugs at effect levels from 18.3 ± 1.8 to 58.2 ± 1.8% inhibition (Table 2 ; Fig. 4A). The AM251 combination with 5-FU at 10 and 25 μM produced an additive effect while, when AM251/5-FU was given at 10/0.5 and 50/0.5 ratio, antagonism was observed on MIA PaCa-2 cells (Table 2). AM251 at 10 and 50 μM dose-dependently increased growth inhibition by 5-FU (Fig. 4A); in these conditions, antagonism, in the presence of favorable DRI values, was also observed (Table 2).
| Concentrations (μM) | Growth inhibition (%) | CI values | DRI values | ||
|---|---|---|---|---|---|
| AM251 | 5-FU | AM251 | 5-FU | ||
| 1 | 0.5 | 18.3 ± 1.8 | 0.53 ± 0.04 ⁎ | 2.98 ± 0.25 | 5.95 ± 0.38 |
| 1 | 1 | 25.7 ± 1.8 | 0.52 ± 0.03 ⁎ | 4.06 ± 0.27 | 3.75 ± 0.19 |
| 1 | 2.5 | 46.9 ± 1.2 | 0.54 ± 0.02 ⁎ | 7.88 ± 0.27 | 2.46 ± 0.06 |
| 1 | 5 | 58.2 ± 1.8 | 0.74 ± 0.03 ⁎ | 10.9 ± 0.58 | 2.46 ± 0.57 |
| 1 | 10 | 79.3 ± 1.5 | 0.80 ± 0.04 | 22.7 ± 1.59 | 1.35 ± 0.07 |
| 1 | 25 | 87.4 ± 0.9 | 1.41 ± 0.06 | 34.4 ± 1.79 | 0.73 ± 0.03 |
| 10 | 0.5 | 36.2 ± 1.5 | 1.86 ± 0.08 ⁎ | 0.56 ± 0.03 | 9.77 ± 0.33 |
| 50 | 0.5 | 84.5 ± 0.66 | 1.79 ± 0.06 ⁎ | 0.57 ± 0.02 | 32.0 ± 0.88 |
- ⁎ P < 0.05.
CI: combination index; DRI: dose reduction index. Data are expressed as means ± S.E. (n = 3). A CI value significantly less than 1 indicates synergy, a CI not significantly different from 1 indicates addition, and a CI significantly higher than 1 indicates antagonism.
4 Discussion
The present study demonstrates in vitro anticancer activity of CB derivatives on the poorly differentiated pancreatic cancer cell line MIA PaCa-2. Our data indicate that such an effect may be mediated via a receptor-independent mechanism and this notion is supported by the following pieces of evidence: (i) both selective CB1 and CB2 receptors agonists and antagonists were cytotoxic; (ii) the absence of CB2 receptors clearly excluded a possible role of these receptors in CBs cytotoxicity; (iii) at nanomolar concentrations, CB1 receptor agonist and antagonist did not significantly change cells viability in serum-free medium, a condition in which non-specific interactions with serum components were avoided; and (iv) at micromolar doses, CB1 receptor did not account for the ACEA-mediated cell death.
Taking into account their lipophilic nature, it is conceivable that CBs may exert cytotoxic effects by crossing cell membrane and directly or indirectly disrupting cellular network with the result of a necrotic and/or apoptotic cell death. To verify this hypothesis, we assessed whether caspase 3/7, two enzymes involved in the effector phase of apoptosis, were a target of CBs. Of note in our study, only the CB1 antagonist AM251 induced a significant time-response effect on caspase 3/7, at a concentration that approximates its IC50 value. Accordingly, internucleosomal degradation of DNA by AM251 was also demonstrated suggesting that apoptosis may be a major contributor to its anticancer action. Furthermore, our results demonstrate that caspase 3/7 inhibition partially reverses the toxicity of AM251; thus it is conceivable that apoptotic cell death may be mediated, at least in part, via a caspase-dependent mechanism.
These results seem to be relevant considering that development of resistant mechanisms for apoptosis confers high survival ability and low drug sensitivity to pancreatic tumor cells [12].
The different apoptotic response of the CBs tested in the current work may be caused by differences in their chemical structures. Indeed, AM251 is a diarylpyrazole derivative homologue to the cyclooxygenase-2 (COX-2) inhibitor celecoxib (Fig. 5 ), which, in turn, promotes apoptosis via activation of caspases 3/7 in the human breast cancer cell line MDA-MB-231 [13]. MIA PaCa-2 is a COX-2-negative cell line [14]; thus, COX-2 inhibition could not account for AM251 cytotoxicity. However, COX-2-inhibitory function is not required for celecoxib-mediated apoptosis which seems to be ascribed to the delayed progression of cells through the G(2)/M phase [15].
The mechanism underlying the cytotoxic action promoted by JWH-015 and ACEA remains to be clarified. However, apart from necrosis, there are several pieces of evidence indicating that cell death can occur in complete absence and independent of caspase activation after cytotoxic drugs administration, for example by autophagy, paraptosis, mitotic catastrophe, and the descriptive model of apoptosis-like and necrosis-like programmed cell death [16]. Some of these mechanisms may be taken into account to interpret the cytotoxic effect of JWH-015 and ACEA on MIA PaCa-2 cells.
To obtain further insights into molecular mechanism of action of AM251, we examined the gene expression profile of MIA PaCa-2 cells after AM251 treatment by microarray analysis. Our data demonstrated that AM251 could affect critical steps of cell proliferation including JAK/STAT and MAPK signaling network as well as cell cycle-related pathways. Despite of the exploratory nature of microarray experiments did not allow weighing the impact of the transcriptional changes in mediating drug response, such a result appears to be in line with the ability of celecoxib to induce apoptosis by modulating several elements controlling cell cycle check points [15, 17].
Finally, we demonstrate that AM251 synergistically enhances the anticancer activity of 5-FU, one of the most commonly used drugs in the treatment of pancreatic carcinoma [18]. Considering the celecoxib-AM251 molecular similarity, these findings support clinical evidences demonstrating the efficacy of the celecoxib/5-FU combination schedule in patients with advanced pancreatic cancer [19].
In conclusion, our data indicate that CB system does not play a relevant role in MIA PaCa-2 cell proliferation and point out AM251 as a prototypical compound for the development of promising diarylpyrazole derivatives useful in fighting pancreatic cancer.
Appendix A Supplementary data
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.febslet.2006.02.024.
Acknowledgements
We thank Drs. Massimiliano Salerno and Erika Melissari for their support to microarray data analysis.
Appendix A A
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.febslet.2006.02.024.
