Background
Pancreatic cancer has a very poor prognosis mainly due to late diagnosis of already advanced tumors, often with metastases to distant organs. Since high resistance to therapy aggravates the treatment outcomes, new efficient treatment modalities and therapy targets are under investigation.
Statins, competitive inhibitors of 3-hydroxyl-methylglutaryl coenzyme A (HMG CoA) reductase, are widely used for treatment of hypercholesterolemia. However, their therapeutic role surpasses the cholesterol lowering capacity, utilizing anti-inflammatory, anti-oxidant and anti-thrombotic actions [
1]. Additionally, several studies suggested the anti-proliferative role of statins in various cancer cell lines, including lung [
2], colorectal [
3] and pancreatic cancer [
4‐
7]. These effects could be partly mediated by the depletion of several important intermediates of cholesterol biosynthesis involved in posttranslational protein prenylation. This process is especially important for modification of small GTPases, such as Ras [
8,
9], which is essential for their translocation from cytoplasm to the cell membrane, affecting thus their cell proliferating activities [
1]
via targeting several important signal transduction pathways [
10‐
12]. The association of activation mutations in the
K-
ras oncogene with pancreatic cancer is well established, being found in more than 90 % of human pancreatic cancers [
13]. We previously reported that most statins protect green fluorescent protein (GFP)-K-Ras from its anchoring to the cell membrane, affecting the signaling pathways and leading to suppression of cancer cell growth in pancreatic cancer cells in vitro [
4].
Heme oxygenase (HMOX), the key enzyme in heme metabolism, catalyzes the degradation of heme to equimolar quantities of CO, free iron and biliverdin, which is subsequently converted to bilirubin [
14]. While the induction of HMOX1 represents a key biological process in adaptive response to cellular stress and displays anti-inflammatory, anti-apoptotic and anti-oxidative actions [
14‐
17], its role in cell proliferation and tumor progression is still controversial [
18,
19]. Some studies suggested that statins can upregulate the
HMOX gene expression in a cell- and species-specific manner [
20‐
24], and they exert some of their protective effects
via this pathway [
21]. However, the upregulation of HMOX1 in pancreatic cancer cells was previously connected to worsened treatment outcome [
25].
The aim of this study was to evaluate anti-proliferative effects of statins with respect to their possible role in modulation of HMOX pathway in pancreatic cancer in vitro. Hemin, a strong HMOX1 inducer [
26], was used a control compound. Further, we investigated the effects of cerivastatin on targeting the GFP-K-Ras protein trafficking, as well as the regulation of invasiveness of pancreatic adenocarcinoma cells in vitro, elucidating the potential involvement of statins in pancreatic cancer therapy.
Methods
Chemicals
Cerivastatin, pravastatin and fluvastatin were purchased from LKT Laboratories, Inc (USA), lovastatin and simvastatin from Santa Cruz Biotechnology (Dallas, TX, USA). Bovine serum albumin (BSA), hemin, reduced nicotinamide adenine dinucleotide (NADPH), sulfosalicylic acid, Dulbecco’s Modified Essential Media (DMEM), and RPMI-1640 were purchased from Sigma-Aldrich (St. Louis, MO, USA). Fetal bovine serum (FBS) and L-glutamine (L-Glu) were purchased from Biosera (Boussens, France), 15-deoxy-Δ-12,14-prostaglandin J2 (PGJ2) was purchased from Merck (Darmstadt, Germany).
Cell culture
For cell culture studies, the following pancreatic cancer cell lines were used: PA-TU-8902 (DSMZ, Braunschweig, Germany), MiaPaCa-2 and BxPC-3 (ATCC, Manassas, VA, USA). All cell lines were maintained and grown in a humidified atmosphere containing 5 % CO
2 at 37 °C. PA-TU-8902 and MiaPaCa-2 were cultured in DMEM supplemented with 10 % FBS, antibiotics and 1 % L-Glu, BxPC-3 in RPMI-1640 supplemented with 10 % FBS, antibiotics and 2 % L-Glu. For all experiments, medium with reduced content of FBS to the final concentration of 0.5 % was used. All statins in the study were used at 12 μM (corresponding to IC
50 of simvastatin for MiaPaCa-2 cells after 24 h incubation [
4]) diluted in methanol (vehicle) and hemin (methemalbumin) was prepared as previously described and used in the final concentration of 30 μM (pH = 7.4) [
26].
Ethical approval for work on cell lines was not required by our Institution.
HMOX RNA interference (RNAi)
Pancreatic cancer cells were transfected with 10 pmol of HMOX1 esiRNA and 10 pmol of HMOX2 esiRNA (Sigma-Aldrich) per 5 x 103 seeded cells using the Lipofectamine RNAiMAX reagent (Life Technologies, Carlsbad, CA, USA) for 24 h in ATB-free DMEM medium. The esiRNA Universal control was used as negative control in all experiments. Data were expressed as % of esiRNA Universal control (Sigma-Aldrich).
Cell proliferation assay
For the cell proliferation assay, cells were seeded into 96 well (5–12.5 x 104 cells per ml according to the cell line) and kept at 37 °C and 5 % CO2. After 24 h, cells were treated with statins or/and hemin, followed by the MTT test (Sigma-Aldrich) as a general cell proliferation assay. As we experienced difficulties with hemin-treated samples using MTT test due to interfering effects of hemin, we further used the more sensitive CellTiter-Glo Luminescent Cell Viability Assay (Promega, Fichburg, WI, USA). Both tests were used according to the manufacturer's instructions. Results were expressed as % of controls.
HMOX activity measurement
Cells in plates were treated with statins and hemin. After 12 h, cells were washed twice with ice-cold phosphate buffer and finally collected into freshly added phosphate buffer and centrifuged. The pellet was resuspended in 150 μl of 0.1 M potassium phosphate buffer (pH = 7.4) and sonicated with an ultrasonic cell disruptor (Model XL2000, Misonics, Farmingdale, NY, USA). The protein concentration was assessed using the DC™ Protein Assay (Bio-Rad Laboratories, Hercules, CA, USA) according to the manufacturer's instruction. A total of 0.15 mg of protein was incubated for 15 min at 37 °C in CO-free septum-sealed vials containing 20 μl of 4.5 mM NADPH as previously described [
27]. The amount of CO generated by HMOX activity was quantified by gas chromatography with a reduction gas analyzer (Peak Laboratories LLC, Mountain View, CA, USA) and calculated as pmol CO/h/mg protein. Five μM PGJ2 was used as a positive control of heme regulation. Results were expressed as % of control.
Western blot analyses
For protein expression analyses, cells were transfected with esiRNA universal control or esiRNA HMOX1/2 as mentioned previously. After 24 h, cells were treated with 30 μm hemin for 20 h. Hemin treatment was used to upregulate HMOX1 protein expression to cumulate detectable levels of HMOX1 protein. Thirty μg of total protein were separated on 12 % polyacrylamide gel and then transferred to nitrocellulose membrane (Bio-Rad Laboratories). After blocking in Tween-PBS with 5 % milk (Sigma-Aldrich) for at least 1 h, membranes were incubated with HMOX1 antibody (1:1000; Thermo Fisher, Rockford, IL, USA), or β-actin (1:1000; Cell Signaling Technology, Danvers, MA, USA) overnight at 4 °C. After washing, membranes were incubated with anti-mouse IgG-HRP (Abcam, Cambridge, UK) for 1 h. Immunocomplexes on the membranes were visualized with ECL Western Blotting Detection Reagents (Cell Signaling Technology).
Real-time PCR analysis of mRNA
HMOX1 expression
Cells grown in plates were treated with statins, hemin or PGJ2. After 4 h, they were washed twice with ice-cold PBS and collected in the lysis buffer. Total cell RNA was isolated using Perfect Pure RNA Cultured Cell Kit (5Prime, Gaithersburg, MD, USA) and cDNA was generated using High Capacity RNA-to-cDNA Master Mix (Life Technologies) according to the manufacturer’s instructions. Real-time PCR for HMOX1 (OMIM *141250) and HMOX2 (OMIM *141251) was performed using the SYBR master mix (Life Technologies) according to the manufacturer’s instructions with optimized primers (Generi Biotech, Hradec Králové, Czech Republic). Results were calculated using the comparative Ct method with HPRT as a house-keeping gene and were expressed as % of control.
Markers of invasiveness
Cells were treated for 12, 24 and 48 h with individual statins. Total RNA was collected and cDNA generated as mentioned above. For real-time qPCR, cDNA corresponding to 10 ng of starting total RNA was diluted with water in 3.6 μl; 0.2 μl of the combined 10 μM forward and reverse primers were added and, finally, 3.8 μl of 2x iTaq Universal SYBR Green Supermix (Bio-Rad Laboratories) was added. The reaction was carried out using the Eco real-time PCR system (Illumina, San Diego, CA, USA) using three-step PCR. The relative mRNA expression levels of osteopontin (secreted phosphoprotein 1, SPP1, OMIM*166490) and sex-determining region Y-related HMG box 2 (SOX2, OMIM*184429) were calculated using the comparative Ct (ΔΔCt) method, with ribosomal phosphoprotein (P0, OMIM*180510) as a reference gene.
Sequences of primers used for real-time PCR: SPP1 forward, AGA CCT GAC ATC CAG TAC CCT, reverse - CAA CGG GGA TGG CCT TGT AT; SOX2 forward - AGG ACC AGC TGG GCT ACC CG, reverse - GCC AAG AGC CAT GCC AGG GG.
Apoptosis evaluation
Apoptosis was quantified using the annexin V-FITC method, which detects phosphatidyl serine externalized in the early phases of apoptosis, in combination with propidium iodide (PI) staining. After exposure to cerivastatin and/or hemin, floating and attached cells were collected, washed with PBS, re-suspended in 100 μl binding buffer and incubated for 20 min at room temperature with 0.3 μl annexin V-FITC (Apronex, Vestec, Czech Republic). PI (10 μg/ml) was added directly before flow cytometry analysis (BD FASC Calibur, BD, Franklin Lakes, NJ, USA). Annexin V positive (An+) and PI negative (PI-) are cells in early apoptosis, An + and/or PI positive (PI+) are cells in late apoptosis or post-apoptotic necrosis.
Reactive oxygen species (ROS) generation
For assessment of ROS generation, dichlordihydrofluorescein diacetate (H2DCFDA) (Life Technologies) was used. After treatment, cells were washed and exposed to 10 μM H2DCFDA in 37 °C for 20 min. Cells were then washed, lysed and the fluorescent signal at 492/520 (Ex/Em) was evaluated in 100 μl aliquots. Total fluorescence was related to protein concentration. Results were expressed as % of control.
Ras protein translocation assay
PA-TU-8902 cells were seeded in dishes with glass bottom 6 h before transfection by pEGFP-KrasWT (GFP – green fluorescent protein, WT-wild type) plasmid prepared as described previously [
4]. Transfection was carried out using FuGene HD according to the manufacturer’s instructions. Cerivastatin (12 μM), pravastatin (12 μM) and hemin (30 μM) were added 12 h post transfection and the cells incubated with the agents for 24 h. Intracellular localization of the GFP-K-Ras protein was visualized by confocal microscopy, using a spinning disk confocal microscope (Olympus, Tokyo, Japan; Andor, Belfast, UK) equipped with solid state laser (488 nm for continual excitation). Emission was collected through a single-band filter (BrightLine® FF01-525 nm, Semrock Inc., NY, USA). The images were obtained and analyzed with the iQ2 software (Andor).
Statistical analysis
All data were expressed as mean ± SD. For normally distributed datasets, one-way ANOVA with post-hoc Holm-Sidak test for multiple comparisons was used for analysis. For non-normally distributed data and small datasets (n ≤ 6), Mann–Whitney rank sum test and Kruskal-Wallis ANOVA with Dunn’s test for multiple comparisons were used. P-values less than 0.05 were considered statistically significant.
All datasets of the results discussed in the manuscript are available on request.
Discussion
Only little progress has been achieved in the treatment of pancreatic cancer over recent decades [
31]. Among multiple tested experimental drugs, statins were shown to display anticancer effects in this malignant disease [
32,
33]. In our study, we investigated a possible relation between statins and the HMOX pathway that play a role in pancreatic carcinogenesis [
25].
Even though HMOX upregulation is associated with beneficial effects for cells, its role in carcinogenesis remains controversial [
19]. HMOX seems to negatively affect the outcome of treatment [
25] and enhance the aggressiveness and progression of pancreatic cancer [
34]. Moreover, pancreatic cancer cells were shown to overexpress HMOX1 compared to normal pancreatic tissue [
25,
34]. On the other hand, statins have been previously shown to upregulate
HMOX gene expression, and some of their protective effects are believed to be mediated
via this pathway [
24]. Together with these HMOX1-inducing effects, statins simultaneously inhibit pancreatic cancer cell proliferation [
4]. In this study, we assessed the overall relationship of particular statins to HMOX regulation in pancreatic cancer cells. We were able to demonstrate no effect of the tested statins on HMOX activity in selected human pancreatic cancer cell lines, despite their remarkable anti-proliferative effects. Moreover, suppression of proliferation by cerivastatin treatment persisted also in
HMOX1- and
HMOX2-silenced cells, indicating that these effects did not depend on the HMOX pathway. Importantly, we noted that
HMOX silencing decreased the cell growth, implying an important role of HMOX in pancreatic cancer cell survival. This is in agreement with previous studies [
25,
34].
To test a possible effect of HMOX on pancreatic carcinogenesis, hemin, a potent HMOX1 inducer [
26], was used in further experiments. To our surprise, a significant decrease in cell proliferation was observed in all tested pancreatic cancer cell lines, despite HMOX induction. Moreover, co-treatment of the cells with hemin and statins increased the anti-proliferative effect of the latter. This is most likely due to hemin-mediated HMOX-independent mechanisms that play a role in cell proliferation and survival. Indeed, we found significant hemin-induced increase in apoptosis, corresponding to considerable increase in ROS production. The same cell growth inhibitory effects were observed even in
HMOX-silenced cells, further suggesting the independence of hemin bioactivity on the HMOX pathway
. Thus, neither statins- nor hemin-dependent suppression of proliferation involved the HMOX system; further, induction of HMOX by hemin did not prevent this response of pancreatic cancer cells to the agents. Nevertheless, our data from HMOX1/2 silencing support pro-carcinogenic role of HMOX in pancreatic cancer, consistent with previous clinical observation [
25].
Despite the fact that hemin was suggested to contribute to increased colon cancer incidence in red meat eaters [
35], other studies shown clear anticancer effects of this compound [
36] supporting our findings.
Similarly as in our previous study [
4] and as discussed recently [
37], we observed remarkable differences in anti-proliferative effects of individual statins, which were dependent also on the cell line used. Both MiaPaCa-2 and PA-TU-8902 cells carry the
K-
ras mutation in codon 12 [
28,
29]. Hamidi and colleagues found that in pancreatic cancer cells, this mutation makes the cells more sensitive to inhibitors of MEK1/2, which is a kinase activated by K-Ras [
29]. To extend our study to pancreatic cancer cells lacking the
K-
ras mutation, we also used BxPC-3 cells carrying a wild-type
K-
ras proto-oncogene and overexpressing cyclooxygenase [
28]. As expected, MiaPaCa-2 cells were the most susceptible and PA-TU-8902 were the most resistant cells to statin treatment in our experiments.
All the statins except pravastatin exerted remarkable growth inhibitory activity in all tested pancreatic cancer cell lines with cerivastatin being the most efficient of these agents. One of the mechanisms possibly involved in anti-carcinogenic effect of cerivastatin might be increased production of ROS, similarly as demonstrated in lymphoma cells [
38]. Interestingly, this phenomenon was much more pronounced in hemin-treated cells and, in particular, in cells exposed to hemin together with cerivastatin.
The K-ras pathway is another possible target of statins. In fact, Gbelcova and colleagues demonstrated that statins, except for pravastatin, prevented the GFP-K-Ras protein from its cell membrane localization in MiaPaCa-2 cells [
4]. We performed a similar experiment in PA-TU-8902 pancreatic cancer cells, and found that while pravastatin did not affect translocation of the K-Ras protein to the cell membrane, cerivastatin significantly prevented GFP-K-Ras from membrane localization. The K-Ras signaling pathway is essential for metastatic lesion formation and tumor invasiveness [
39]. Interestingly, cerivastatin but not pravastatin treatment of PA-TU-8902 cells significantly decreased the expression of
SPP1 and
SOX2, factors with important role in cancer metastasis and aggressiveness [
40,
41]. Meta-analysis of 11 studies revealed that patients with pancreatic cancer have elevated serum levels of SPP1 [
42]. Similarly,
SOX2 over-expression promotes self-renewal and de-differentiation of pancreatic cancer cells [
43]. In our experiments, both markers were downregulated in pancreatic cancer cells exposed to cerivastatin, pointing to a lowering effect of the statin on the metastatic potential of the cells.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
KV performed the most of cell culture studies, qPCR experiments and drafted the manuscript; SB performed flow cytometry studies; HG performed confocal microscopy studies; LM performed heme oxygenase activity and RNAi experiments; JN was involved in study design, flow cytometry experiments and assisted with drafting of the manuscript; RG contributed to the study design, participated in qPCR experiments, and assisted with drafting of the manuscript; TR contributed to confocal microscopy experiments and assisted with drafting of the manuscript; LV designed the study, supervised all the experiments and assisted with drafting of the manuscript. All authors read and approved the final manuscript.