Background
The mammalian Pim family of serine/threonine-specific kinases consists of three highly homologous proteins, Pim-1, Pim-2 and Pim-3, whose functions and expression patterns are partially overlapping [
1,
2]. Crystallization studies have revealed that Pim kinases constitutively reside in an active conformation [
3], so that their activities are mainly regulated at the level of expression. In hematopoietic cells
pim expression is transiently induced by a number of cytokines including several types of interleukins (IL; [
4‐
6]). When overexpressed in mice, Pim kinases are oncogenic and can enhance lymphomagenesis, especially in collaboration with other oncoproteins such as Myc family members, Bcl-2 or Gfi-1 [
7‐
10]. Upregulated expression levels for Pim kinases have been observed not only in human leukemias and lymphomas [
11], but also in several types of solid tumors such as prostate, colon, oral, hepatic and pancreatic cancers (reviewed in [
12,
13]).
There are many ways how Pim kinases contribute to tumorigenesis by promoting proliferation and preventing apoptosis (reviewed in [
13,
14]). In hematopoietic cells, we have previously shown that Pim-1 stimulates activities of several transcription factors including c-Myb [
15], NFATc1 [
16] and the RUNX family proteins [
17], and thereby enhances production of cytokines and other survival factors. In addition, all Pim family members inhibit apoptosis by phosphorylating and thereby inactivating the pro-apoptotic Bad protein [
18‐
20]. However, the physiological role of Pim kinases in development of solid tumors has remained fairly elusive.
The emerging importance of Pim kinases in human tumorigenesis has raised growing interest to develop small molecule inhibitors for them. Several different classes of Pim inhibitors have recently been reported (reviewed in [
13]), but only a few of them have been tested in cell-based assays or animal models to demonstrate anticancer activity [
21‐
27]. In addition, only a few of them are effective against all Pim family kinases [
21,
25‐
27]. Due to functional redundancy [
2], simultaneous targeting of all Pim kinases can be of advantage in treating cancer patients. Yet no severe side effects are expected, since mice lacking all three Pim family members are just slightly deficient in their growth responses, but otherwise viable and fertile with a normal life span [
2].
In this study, we have further analysed the cellular effects of 1,10-dihydropyrrolo[2,3-
a]carbazole-3-carbaldehyde (DHPCC-9) that we recently identified as a potent and selective inhibitor against all three Pim family kinases [
21]. We now demonstrate that this inhibitor can efficiently block several cellular functions of Pim kinases. Furthermore, using this inhibitor along with RNA interference or protein overexpression, we have been able to reveal an as yet unrecognized role for all Pim family kinases in promoting migration and invasion of adherent cancer cells.
Methods
Cell lines and culture conditions
The murine IL-3-dependent myeloid FDCP1 cell lines and the human head and neck squamous cell carcinoma cell line UT-SCC-12 have been previously described [
28,
29]. FDCP1 cell lines and the human androgen-independent prostate epithelial adenocarcinoma cell line PC-3 (American Type Culture Collection) were maintained in RPMI-1640 medium, while UT-SCC-12 cells were cultured in DMEM medium with 1% non essential amino acids. All media were supplemented with 10% fetal bovine serum. 10% WEHI-conditioned medium was used as the source of IL-3 for FDCP1 cell lines.
Cell viability assays
For MTT assays, cell cultures were incubated for 4 h with 0,5 mg/ml 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) reagent (Sigma-Aldrich). Formazan crystals formed during the assays were dissolved by acidic isopropanol. Optical densities were determined by spectrophotometry (Multiscan MCC/340, Labsystems). Alternatively, cells were stained with Trypan blue (Sigma-Aldrich) and live cells excluding the dye were counted.
Western blotting
Cell pellets were resuspended and lysed in NP-40 lysis buffer (50 mM Tris, pH 7.5, 10% glycerol, 100 mM NaCl, 1 mM EDTA, 1% Nonidet P-40 plus protease and phosphatase inhibitors). After clearing the lysates, protein concentrations were measured with DC Protein Assay (Bio-Rad). Twenty to hundred microgram aliquots of protein were separated by SDS-PAGE, immobilized onto PVDF-membrane (Millipore) and incubated with anti-Pim-1 (12H8; Santa Cruz), anti-Pim-2 (D1D2; Cell Signaling Technology for human protein or HPA000285; Atlas Antibodies for murine protein), anti-Pim-3 (D17C9; Cell Signaling Technology), anti-V5 (Invitrogen), anti-Flag or anti-GAPDH (Sigma-Aldrich) antibodies. Chemiluminescence reactions were generated by either Amersham™ ECL Plus (GE Healthcare) or Pierce® ECL (Thermo Scientific) reagents. The signal intensities were quantified by MCID M5+ Image Analyzer (InterFocus, UK).
Cell-based phosphorylation assays
FDCP1 derivatives were transiently transfected with 10 μg of the GST-Bad expression vector (pEBG-mBad; [
19]) using the GenePulser II electroporator (Bio-Rad). Cells were lysed in buffer containing 10 mM Tris, pH 7.5, 150 mM NaCl, 0,5 mM EDTA, 1% Triton X-100, 10% glycerol and phosphatase inhibitors. Cleared lysates containing 60 μg aliquots of protein were used to purify GST-Bad protein with glutathione Sepharose beads (GE Healthcare) at 4°C. The precipitates were fractionated by SDS-PAGE and subjected to Western blotting with anti-Bad and anti-phospho-Bad (Ser
112) antibodies (Cell Signaling Technology).
Lipofections
For RNA interference, PC-3 cells were transfected with short interfering RNAs (siRNAs) and Oligofectamine™ (Invitrogen). Non-targeting control siRNA (D-001810-01-20) or specific siRNAs targeting either pim-1 or pim-2 (Dharmacon) were used at 100 nM concentration. These ON-TARGETplus siRNAs have been specifically designed to reduce off-target effects that may lead to toxicity or false phenotypes.
For overexpression, Fugene transfection reagents (Roche) were used to transfect PC-3 cells with 1 μg of the pcDNA3.1/V5-HisC vector or its derivatives expressing human pim genes (N.M. Santio, M. Varjosalo, J. Taipale and P.J. Koskinen, manuscript in preparation), or 0,25 μg of the pBJ5 vector or pBJ5-NFATc1-FLAG obtained from S.N. Ho (Stanford University, CA).
Wound Healing assays
Wound healing assays were performed on 24-well plates either manually or automatically. In manually performed assays, different concentrations of DHPCC-9 were used, while DMSO concentration was maintained at 0,1%. Scratch wounds were made with a sterile 10 or 200 μl pipette tip. Photographs were taken using the Zeiss Stereo Lumar-V12 microscope with the AxioVision Rel.4.8 software. Percentages of wound healing were calculated by the ImageJ software (Wayne Rasband, NIH, USA) and the approximate edges of the wounds were manually marked to figures with straight lines.
For automatic assays with the WoundMaker™ and the IncuCyte™ systems (Essen Instruments), the scratch wounds were made with 10 μl sterile pipette tips, after which fresh culture medium containing 10% or no serum was added along with either DMSO or 10 μM DHPCC-9. The IncuCyte™ Scratch Wound software was used to capture and analyse the pictures. After acquisition, images were combined with ImageJ into a QuickTime movie file.
Boyden chamber invasion assays
Cell culture invasion inserts of 8 μm pore size (BD Biosciences) were coated with Matrigel (100 μg/cm2; BD Biosciences) and incubated for 24 h. The assays were initiated by placing 50 000 cells in DMEM supplemented with 1% BSA together with either 10 μM DHPCC-9 or 0,1% DMSO. Conditioned medium from confluent MG-63 human osteosarcoma cells was used as a chemoattractant to stimulate movement of cells through the in vitro basement membranes. Cells were incubated for 72 h, after which insert membranes were fixed for 20 min in 4% paraformaldehyde in PBS and stained with Mayer's haematoxylin (Zymed) for 4 h. Then membranes were washed with PBS and cut from the inserts. Cells on the upper surface of the membrane were wiped off and membranes were mounted with glycerol and PBS (9:1, Merck KGaA). Invaded cells on the lower surfaces of the membranes were counted.
Statistical analyses
The statistical significance of data was determined by pairwise comparisons between control samples and treated samples by using Student's t-test (Paired Two Sample for Means). Results were interpreted as highly significant*** (p < 0.001), significant** (p < 0.01), weakly significant* (p < 0.05) or not significant
ns
(p > 0.05). IC50 values of DHPCC-9 in FDCP-1 cells were determined using nonlinear regression fitting with the GraphPad Prism v.5.0. Error bars in all graphs represent SD values.
Discussion
In this study, we have characterized the biological effects of 1,10-dihydropyrrolo[2,3-a]carbazole-3-carbaldehyde (DHPCC-9) and demonstrated that it is an efficient cell-permeable inhibitor targeting all human or murine Pim family kinases. We have shown that DHPCC-9 abrogates the anti-apoptotic effects of Pim-1 in cytokine-deprived FDCP1 myeloid cells, while it does not display general cytotoxicity at the micromolar concentrations used. DHPCC-9 treatment also inhibits intracellular phosphorylation of Pim substrates such as Bad. Furthermore, we have been able to use DHPCC-9 as a tool to reveal a novel function for Pim kinases in regulation of motility of adherent cancer cells. Treatment with DHPCC-9 significantly slows down migration of both PC-3 prostate cancer cells and UT-SCC-12A squamocellular carcinoma cells, but does not significantly affect metabolic activity or viability of these cells or their levels of Pim expression. In addition, DHPCC-9 efficiently inhibits invasion of PC-3 cells through matrigel.
The observed effects of DHPCC-9 are most likely dependent on its ability to inhibit Pim kinase activity, since similar results were obtained by silencing Pim expression by RNA interference. In addition, overexpression of any Pim family member has opposite effects by enhancing cancer cell motility, providing further proof that Pim kinases are potent regulators of cellular processes involved in migration and/or invasion.
In all of our assays, the effects of DHPCC-9 were more pronounced than those of Pim-specific siRNAs, most likely due to the longer half-life and better cellular penetrance of the inhibitor. Indeed, when the stability of DHPCC-9 dissolved in the growth medium of PC-3 cells was analysed by liquid chromatography, no detectable degradation of the compound was observed within the time-periods used in our assays (unpublished data). By contrast, siRNAs affect only a portion of cells depending on transfection efficiency and their transient effects are rapidly diluted within the few days after transfection. Here it should also be noted that while DHPCC-9 is able to target all the three Pim kinases endogenously expressed in PC-3 cells, Pim-3 remains active in cells treated with the combination of siRNAs against Pim-1 and Pim-2. Yet in PC-3 cells the expression levels for Pim-3 have been reported to be significantly lower than for the two other family members [
27]. Thus, even though all Pim kinases may be equally capable of enhancing cell motility, the magnitude of their redundant effects is likely to depend on their dosage in each cell type.
An intriguing feature of Pim family kinases is that they are constitutively active with a unique hinge region containing a proline residue, which allows only one hydrogen bond to be formed with ATP [
3]. According to crystallization studies [
21], DHPCC-9 does not bind to the hinge region as some other ATP-mimetic Pim inhibitors do [
33,
34]. By contrast, it competes with ATP for binding to the conserved lysine residue corresponding to Lys67 in Pim-1 [
21]. This residue is known to be critical for Pim kinase activity, since its mutation e.g. to methionine results in production of a kinase-deficient protein [
16].
Under
in vitro conditions, we have previously demonstrated that DHPCC-9 is highly potent against Pim-1, Pim-2 and Pim-3 kinases with very low IC
50 values of 12, 51 and 10 nM, respectively [
21]. It is also highly selective, since with the 10 μM concentration used in most of our cellular assays, the
in vitro activities of Pim-1, Pim-2 and Pim-3 kinases are inhibited by 98, 93 and 99%, respectively, while all the other 88 kinases tested remain fairly active ([
21] and unpublished data). Yet we cannot fully exclude the possibility that some of the observed effects of DHPCC-9 are enhanced by its moderate ability to inhibit also other kinases. Interestingly, DHPCC-9 seems to target Pim-1 and Pim-3 slightly more efficiently than Pim-2. This selectivity is in line with similar observations from several other Pim inhibitors, as recently discussed [
13]. Even though Pim-1 and Pim-3 are more closely related to each other than to Pim-2, structural models are unable to explain the observed differences in their sensitivities.
In cell-based assays, the activities of inhibitors are not expected to be as high as under
in vitro conditions with purified compounds. Thus, the micromolar IC
50 values observed in FD/Neo (6.0 μM) and FD/Pim44 (4.7 μM) cells can be considered promising, especially since they were obtained using complete medium containing 10% serum. Cytotoxicity assays with some other Pim inhibitors have been carried out in the presence of lower or no serum levels [
22,
27], which probably has decreased the viability of the cells and also enhanced bioavailability of the inhibitors and consequently lowered their IC
50 values. Indeed, more pronounced effects were observed also in our assays with DHPCC-9, when cells were grown in the absence of serum.
Yet it is likely that the efficiency of the Pim inhibitor can be further improved by using DHPCC-9 as a scaffold for production of additional, more potent derivatives, which could be useful not only as research tools, but also as lead compounds in development of drugs against Pim-overexpressing tumors. Since DHPCC-9 has been shown to be highly efficient in reducing the motility of Pim-overexpressing cancer cells, derivatives of DHPCC-9 might be able to prevent tumor metastasis and/or angiogenesis. Moreover, since we and others have recently shown that Pim kinases are involved in development of resistance against radiation therapy or chemotherapy [
29,
35], combinations of Pim inhibitors together with other anticancer therapy methods are expected to lead to most efficient therapeutic approaches.
Even though Pim kinases have been implicated to have prognostic roles in several types of solid cancer, there is still controversy in the literature on whether or not high levels of Pim expression are of disadvantage for prostate cancer patients (reviewed in [
12,
13]). This may be partly due to heterogeneity of the samples and to the fact that in none of the studies published so far have expression levels for all three Pim kinases been analysed in parallel. It is also clear that overexpressed Pim kinases alone are unable to transform cells, but require collaboration e.g. with Myc oncoproteins. Interestingly in this regard, coexpression of Pim-1 and c-Myc in human prostate tumors has recently been associated with higher Gleason grades than overexpression of either one alone, suggesting that these oncoproteins synergize to induce advanced prostate carcinoma [
36].
While our work was in progress, silencing of Pim-3 was reported to reduce endothelial cell spreading, migration and vascular tube formation [
37], providing further support to our hypothesis that Pim kinases can stimulate metastatic and/or angiogenic potential of cancerous cells. In addition, Pim-1 but not Pim-2 was shown to regulate homing and migration of bone marrow cells, possibly via phosphorylation-mediated modification of CXCR4 expression on cell surface [
38]. Yet the exact substrates and signalling pathways that all three Pim kinases regulate to enhance motility of adherent cancer cells remain to be elucidated. Interestingly, the NFATc transcription factors that we have previously identified as Pim targets [
16] have recently been implicated in tumor cell migration and invasion as well as tumor angiogenesis (reviewed in [
39]). Constitutively active NFATc isoforms have been shown to promote induction and progression of both hematological malignancies and solid tumors by driving synthesis and secretion of pro-angiogenic factors (e.g. cyclo-oxygenase 2 and tissue factor) as well as factors promoting cell motility (e.g. lysophosphatic acid and prostaglandin E2). Thus, these NFATc-dependent effects are expected to be enhanced in tumor cells overexpressing Pim kinases. Indeed, our data indicate that Pim inhibitors can block the pro-migratory effects of NFATc factors in prostate cancer cells, suggesting that regulation of NFATc activity may be one of the mechanisms how Pim kinases promote cancer cell motility.
Acknowledgements
We thank P. Goekjean for providing the compounds for initial analyses and for bringing the authors together, R. Grenman for the UT-SCC-12 cell line, S.N. Ho for the NFATc construct and R. Andersen and A. Viitanen for technical assistance. We also thank the Cell Imaging Core of Turku Centre for Biotechnology for assistance in flow cytometry and microscopy.
Most parts of this study were supported by the Academy of Finland (grants 111820 and 121533 to PJK). Production of DHPCC-9 was supported by European Union Prokinase Research Consortium and CNRS Valorisation (MP, PM and FA). The experimental work was also supported by the Drug Discovery Graduate School (NMS, RLV, SSV), Turku University Foundation (NMS), the Orion-Farmos Research Foundation (RLV), the Turku University Hospital Fund (RLV), the Cancer Society of Southwestern Finland (RLV) ja the Paulo Foundation (RLV).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
NMS designed and carried out the migration and invasion analyses and drafted the manuscript. RLV analysed inhibitory activities of various compounds in cell-based assays and helped to draft the manuscript. EMR helped in phosphorylation and wound healing assays, JAS in image analyses and SSV in invasion assays. FA, PM and MP synthesized the DHPCC-9 inhibitor, tested its stability and provided sufficient amounts of it for the biological experiments. PJK conceived and coordinated the study, designed the cell-based assays to recognize cellular inhibitors for Pim kinases and edited the manuscript. All authors read and approved the final manuscript.