Introduction
Integrin-linked kinase (ILK), an intracellular serine/threonine kinase, is a key signaling molecule expressed in most, if not all, tissues, with high levels of expression in normal pancreatic, cardiac and skeletal muscle tissues. Through interactions with a diverse range of proteins including adapters such as particularly interesting Cys-His-rich protein (PINCH), calponin homology-containing ILK-binding protein (CH-ILKBP), affixin and paxillin, kinases such as integrin-linked kinase-associated serine/threonine phosphatase 2C (ILKAP), protein kinase B (AKT) and phosphoinositide-dependent kinase 1 (PDK-1), and transmembrane receptors such as β1 and β3 integrins [
1], ILK is thought to play a key role in integrin and growth factor receptor related signaling cascades [
2,
3]. For example, ILK acts as a scaffold protein to allow for protein-complex formations connecting extracellular integrin signals to intracellular actin cytoskeleton rearrangements through direct interaction with the cytoplasmic domain of β1 integrin [
4]. Cell extracellular matrix (ECM) adhesion complexes influence a vast number of cellular processes including cellular morphology, migration, proliferation, survival, and differentiation. Activation of downstream targets of ILK such as AKT [
5], glycogen synthase kinase 3 (GSK-3) [
6], myosin light chain (MLC) [
7], affixin [
8] and the cytoplasmic domain of β1 integrin [
9], is associated with signaling cascades known to regulate transcription of genes involved in a diverse range of functions including: cell survival, cell cycle progression, cell adhesion and spreading, focal adhesion plaque formation, ECM modification, cell motility, and contractility [
1,
10].
Increased ILK expression and activity is found in association with many cancer types including: breast, brain, prostate, pancreatic, colon, gastric, ovarian, and malignant melanomas [
4,
11‐
16]. Further, there is mounting experimental evidence indicating that ILK plays a pivotal role in many processes associated with tumorigenesis. Enforced over-expression of ILK in immortalized rat intestinal epithelial cells induces epithelial to mesenchymal transition (EMT) and a transformed tumorigenic phenotype that is, in part, linked to ILK-dependent inhibition of E-cadherin expression and increased nuclear translocation of β catenin. Over-expression and constitutive activation of ILK leads to dysregulated growth and suppression of apoptosis and anoikis [
17,
18]. With specific respect to breast cancer, over-expression of ILK in mammary cells stimulates anchorage-independent cell growth, cell cycle progression, and increased cyclin D and A expression
in vitro [
2,
19]. Furthermore, mammary epithelial cells over-expressing ILK exhibit hyperplasia and tumor formation
in vivo. [
4]. Further evidence has indicated ILK might play a key role in VEGF-mediated endothelial activation and angiogenesis [
4,
20].
Targeted inhibition of ILK in cancer cells by various strategies can also lead to suppression of the AKT signaling pathway, inhibition of cell cycle progression, reduced vascular endothelial growth factor (VEGF) secretion
in vitro, and reduced tumor growth
in vivo [
21]. A number of pharmaceutically viable small-molecule inhibitors of ILK have been developed and partially characterized. From the K15792 class of the pharmacophor family [
22], some of these inhibitors were shown to effectively inhibit cancer cell survival, growth [
23] and invasion [
24], and induce apoptosis and cell-cycle arrest
in vitro [
25], as well as inhibit tumor growth and angiogenesis
in vivo [
20]. Interestingly, the most promising ILK inhibitor, QLT0267 (267), while capable of eliciting pleiotropic effects in xenograft models of glioma, was unfortunately shown to only delay, but not prevent, tumor growth
in vivo, even at doses as high as 200 mg/kg [
2,
23]. Based on these findings, we speculate that optimal therapeutic effects of 267 will only be realized using a combination therapeutic strategy.
Here we demonstrate on the basis of a cell viability assessment determined using multiple breast cancer cell lines that 267 in combination docetaxel (Dt) interacted in a synergistic manner (increased therapeutic benefit over single agents as assessed by the median effect methodology developed by Chou and Talalay [
26]). Experimentations aimed to identify underlying molecular mechanisms and additional drug-drug interactions using multiple endpoint analyses, revealed in breast cancer cells expressing low levels of Her2, beneficial drug-drug interactions on the basis of endpoints measuring AKT phosphorylation and F-actin cytoarchitecture. Using an orthotopic model of breast cancer (low Her2), 267/Dt combinations were found to exert enhanced therapeutic activity, as demonstrated by significantly reduced tumor growth and extended survival in mice treated with the combination compared to the single agents.
Materials and methods
Chemicals
Cisplatin, doxorubicin, paclitaxil, Dt, vinorelbine, and trastuzumab (Tz) were obtained from the British Columbia Cancer Agency Pharmacy (Vancouver, BC, Canada) and 267 was a generous gift from QLT Inc (Vancouver, BC, Canada). All other chemicals, unless specified, were purchased from Sigma Chemical Company (Oakville, Ontario, Canada). Dt was reconstituted in 13% ethanol for a final concentration of 10 mg/ml and Tz (Hoffman-La Roche, Mississauga, Ontario, Canada) was reconstituted in PBS at a stock concentration of 21 mg/ml.
Cell-lines and culture
MCF-7, KPL-4, BT-474, MDA MB/468 and SKBR3 cells were purchased from American Type Culture Collection (Manassas, VA, USA). MDA-MB-435 (LCC6) estrogen receptor-negative breast cancer cells [
27] and MCF-7
Her2 cells were generously donated. LCC6
Her2 cells, previously described by our group [
28], were generated by the stable transfection (neomycin selection using G418) of plasmid DNA containing the
Her2 gene driven by the cytomegalovirus promoter. LCC6 cells were stably transfected using a lenti-virus system with the luciferase gene and green fluorescent protein (GFP). Cells were sorted by FLOW cytometry for GFP expression and selected cells were used in the following experiments. Sorted cells exhibited similar
in vitro and
in vivo growth rates as the parental LCC6 cell line. Additionally LCC6
luc and parental LCC6 were equally sensitive to Dt.
The breast cancer origin of the LCC6 parental cell line, MDA-MB-435, is controversial. Based on studies of Ross and colleagues [
29] and Rae and colleagues [
30] it has been suggested that the MDA-MB-435 cell line is of a melanoma origin. However, Sellappan and colleagues [
31] have been able to demonstrate that MDA-MB-435 cells can be induced to express breast differentiation-specific proteins and secrete milk lipids. Further, more recent studies of Neve and colleagues [
32] have demonstrated that the MDA-MB-435 cell line shares many molecular features with breast cancer cell lines of breast epithelium origin. In studies from our laboratory [
28] using a LCC6 cell line permanently transfected with the
Her2 gene (LCC6
Her2 cell line), we have been able to demonstrate that the Her2-positive variant exhibit enhanced survival under stress, overproduction of VEGF, activation of nuclear factor (NF) κB and
in vivo sensitivity to Tz (aka Herceptin™; Hoffman-La Roche, Mississauga, Ontario, Canada); results that are consistent with what is known about Her2-positive breast cancer models. Thus, we believe it is justifiable to use these cells as a model breast cancer cell line; particularly when the results obtained using this cell line are confirmed with other breast cancer cell lines.
LCC6, LCC6Her2, LCC6luc, KPL-4, BT-474, MDA MB/468, MCF-7 and MCF-7Her2 cells were maintained in Dulbecco's modified eagle's medium (DMEM)/high glucose supplemented with L-glutamine (2 mmol/L; DMEM and L-glutamine from Stem Cell Technologies, Vancouver, BC, Canada) 5 mM penicillin/streptomycin, and 10% FBS (Hyclone, Logan, UT, USA). SKBR3 cells were maintained in McCoy's 5a medium (fStem Cell Technologies, Vancouver, BC, Canada) supplemented with L-glutamine, 5 mM penicillin/streptomycin, and 10% FBS. All cells were maintained at 37°C and 5% carbon dioxide in a humidified atmosphere.
Cell viability assays
Metabolic activity (measure of cell viability) of breast cancer cell lines incubated in the presence of various therapeutic agents was determined using Alamar Blue® assays (Medicorp Inc. Montreal, QU, Canada) according to the manufacturer's suggestions. Briefly, 6000 cells/well seeded in triplicate onto 96-well flat-bottom tissue culture plates (Techno Plastic Products AG, Trasadingen, Switzerland) were allowed to adhere to the substratum for 24 hours under normal growth conditions (37°C and 5% carbon dioxide in a humidified atmosphere). Serial dilutions of individual drugs, 267/drug combinations and vehicle controls diluted in appropriate cell culture medium were then added to the wells and cells were grown for an additional 72 hours. To assess cell viability, cells were then incubated with 10% resazurin solution for four hours at 37°C and fluorescence was measured at 560/590 nm using an Optima fluorescence plate reader (BMG Labtech, Durham, NC, USA). Relative fluorescence determined from drug-treated cells was normalized to fluorescence determined from control cells (cells grown in presence of appropriate vehicle control alone) and data is shown as percentage relative cell viability compared with vehicle-treated control cells (100% viability, highest fluorescence). Background fluorescence was subtracted from all samples and results of experiments conducted in triplicate are indicated (average ± standard deviation).
Drug combination effects – median effect principle
To determine whether various 267/drug combinations had resulted in synergistic, antagonist, or additive effects, the median effect principle (MEP) method of Chou and Talalay was used to determine combination index (CI) values [
26,
33,
34]. Briefly, the MEP method is used to describe and understand the relationship between a measured response within a population of cells (fraction affected (f
a) versus the fraction unaffected (f
u)) and the fraction of the dose (D) required to achieve an effect level of 50% and is represented by the formula:
where Dm is the dose required to achieve a 50% effect level and m is a coefficient indicating the sigmoidicity of the dose-effect curve. The right side of the equation [(D/Dm)m] represents the dose, and the left side of the equation [fa/fu] represents the effect of the interaction. The CI can be calculated at any effect level and the effect used can be derived on the basis of different endpoints (e.g. cell viability, inhibition of VEGF secretion, etc.). If CI is equal to one then the combination interactions result in additive effects, if the CI is less than one the combination interactions are considered synergistic, and if the CI is greater than one the combination interactions are considered antagonistic.
To determine CI values, the commercially available program CalcuSyn (Biosoft Ferguson, MO, USA) was used to calculate CI values for a broad range of effect levels and, on the basis of this analysis, Fa versus CI plots were generated. CI values were then used to estimate the dose reduction index (DRI) for combination of drugs. The DRI estimates the extent to which the dose of one or more agents in the combination can be reduced to achieve effect levels that are comparable with those achieved with single agents. Drug combinations that acted synergistically can be identified as those that exhibited significant dose reduction values (i.e. a given measured effect will be observed at dose(s) significantly lower than expected based on single agent activities).
VEGF expression
To determine whether a specified treatment influenced VEGF expression, ELISA assays using Quantikine Human VEGF Immunoassay kits (R&D Systems, Minneapolis, MN, USA) were conducted according to manufacturer's suggestions. Briefly, 6000 cells were seeded onto 96-well tissue culture plates and allowed to adhere for 24 hours. Cells were then grown in the presence of single agents or combinations of drugs for 72 hours (as described above). The experiments were completed in triplicate and repeated at least two times. Supernatants were collected, combined, and then assayed for the presence of secreted VEGF (specific for recombinant human VEGF165 and recombinant human VEGF121) using the Optima fluorescence plate reader (BMG Labtech, Durham NC, USA). Results were normalized to total protein found in supernatant and compared with standard curves determined using VEGF standards provided in the kit. This assay accurately measures VEGF levels between 9 pg/ml and 2000 pg/ml.
Western blot analysis
Total protein lysates were prepared from cells incubated in the presence of single drug, the drug combinations or vehicle controls. Briefly, cells were rinsed with PBS, harvested from plates with trypsin, and centrifuged at 1500 × g for five minutes. Cell pellets were then re-suspended in lysis buffer (150 mmol/L sodium chloride, 1% NP40, 0.5% sodium deoxycholate, 2.5 mmol/L EDTA, 0.1% SDS), Mini protease inhibitor cocktail tablets (Roche Diagnostics, Mannheim, Germany), sheared using 25-gauge needles, incubated on ice for 30 minutes, and finally centrifuged at 10,000 × g for 10 minutes to remove insoluble material. Protein concentrations were determined from supernatant using the Bradford Method and approximately 75 μg of total protein from each sample were denatured in loading buffer (Invitrogen, Burlington, ON, Canada) by boiling for 10 minutes and loaded onto 10% SDS-PAGE. Proteins separated by electrophoresis were transferred to Nitrocellulose membrane (Millipore, Bedford, MA, USA) and blocked for one hour at room temperature in Odyssey blocking buffer (Licor Biosciences, Lincoln, NB, USA). Membranes were incubated at 4°C overnight in Odyssey blocking buffer containing polyclonal anti-ILK, anti-AKT, anti-P-AKT or anti-Her2 antibodies (1:1000 dilution; Cell Signalling Technology, Beverly, MA, USA). Membranes were then washed three times for five minutes with PBS-Tween (1% v/v) and incubated with either anti-rabbit or mouse IRDYE (green) (Rockland, Gilbertsville, PA, USA) or anti-rabbit Alexa 680 (red) (Invitrogen, Molecular probes, Burlington, ON, Canada) (1:10,000) for one hour at room temperature and signals were detected and quantified using the Odyssey Infrared Detection System and associated software (Odyssey v1.2; Licor, Lincoln, NB, USA). Background and input variation between samples were corrected using signal intensities for negative control pixel noise and actin band intensities, respectively. Data were expressed as mean values ± standard deviation and parametric analysis was performed using an unpaired Student t-test.
Immunofluorescence analysis
Cells grown on coverslips were rinsed with PBS (pH 7.4), fixed using 2.5% paraformaldehyde (w/v) in PBS for 20 minutes at room temperature and permeabilized using 0.5%Triton X-100 (v/v) in PBS for five minutes at room temperature. Coverslips were then washed three times with PBS and incubated for one hour in 2% BSA (w/v) in PBS to block non-specific binding, washed three times in PBS, and then incubated with phalloidin conjugated to Texas red (1:500) (Molecular Probes, Eugene, OR, USA) for 20 minutes at room temperature. Nuclei were stained using Hoechst nuclear stain (10 mg/ml) (Molecular Probes, Eugene, OR, USA) (1:1000) for 15 minutes at room temperature. Coverslips were rinsed once with double distilled water and mounted to microscope slides using a 9:1 solution of glycerol and PBS (Air Products & Chemicals, Inc., Allentown, PA, USA). Images were viewed and captured using a Leica CTR-mic UV fluorescent microscope (Wetzlar, Germany) and a DC100 digital camera with Open Lab software (Improvision, Lexington, MA, USA).
Tumor xenografts
All animal studies were conducted in accordance with institutional (University of British Columbia) guidelines for humane animal treatment and according to the current guidelines of the Canadian Council of Animal Care. Mice were maintained at 22°C in a 12-hour light and dark cycle with ad libitum access to water and food. Two million LCC6luc cells were injected into the mammary fat pad of female NCr nude mice (Taconic, Oxnard, CA, USA) in a volume of 50 μL using a 28-gauge needle. Tumor growth was monitored using an IVIS 200 non-invasive imaging system (Xenogen, Caliper Life Sciences, MA, USA), and manually using callipers when tumor dimensions exceeded 3 mm in length and width. Tumor volume (mm3) estimated from length and width measurements were calculated according to the equation length times width squared divided by two with the length (mm) being the longer axis of the tumor. Animal body weights were recorded every Monday and Friday.
In vivoimaging system
Imaging was performed once every seven days to monitor tumor progression. LCC6luc tumor-bearing mice were injected intraperitoneally with 500 μl D-luciferin (15 mg/ml) (Xenogen Corp., Alameda, CA, USA). Mice were anesthetized using isoflurane and twenty minutes post intraperitoneally injection mice were imaged. Photographic and luminescence images were taken at exposure times of one, two, and five second(s) and Xenogen IVIS® software was used to quantify non-saturated bioluminescence in regions of interest (ROI). Light emission between 5.3067 × 106 and 2.2179 × 109 was determined to contain tumor tissue while emissions below this range were considered as background. Bioluminescence was quantified as photons/second/cm2/steradian for each ROI.
Statistical analysis
All statistical data was collected using GraphPad InStat (San Diego, CA, USA). One-way analysis of variance was performed using standard error of the mean, mean and n and a Tukey-Kramer Multiple Comparisons Test was used as the post hoc test.
Discussion
Although it is understood that ILK is an important therapeutic target in cancer, the data summarized here (Figure
8) and elsewhere suggest that an ILK inhibitor such as 267 given alone will not achieve much more than a delay in tumor progression. Lack of potent single-agent activity, when using
in vivo tumor growth as an efficacy measure, lends support to the belief that ILK inhibitors must be developed in the context of other therapeutics. A similar trend was exemplified by treatment regiments incorporating Tz (Herceptin™), a therapy that targets Her2-expressing tumors. Tz as a single agent exhibits little significant activity, but when used in a combination setting it has proved to be of significant therapeutic value [
38]. The studies described here, focused on identifying agents that would work synergistically with QLT0267. We used cell-based screening assays in order to assess whether drugs commonly used for breast cancer could be combined with 267 to achieve better then expected therapeutic results. For these studies a fixed-drug ratio experimental design was used where drug-drug interactions were determined using at least three different drug-drug ratios applied over a broad range of effective doses (Table
2). We show for the first time that combination of 267/Dt appeared to interact in a manner that results in synergy. Drug-drug interactions were measured by use of the median effect method of Chou and Talalay [
26] and were initially determined on the basis of a therapeutic endpoint measuring metabolic activity (Alamar Blue assay). Synergy was observed over a broad range of effective dose and was measured in five out of six breast cancer cell lines tested (Figure
2), regardless of Her2 status. Although limited to results obtained with the two cell lines used for the broad combination screen (LCC6 and LCC6
Her2) it is interesting to note that the 267/Dt combination was synergistic while combinations of 267 with paclitaxel and vinorelbine appeared antagonistic. This would suggest that the mechanism(s) promoting synergy may not involve microtubules in general. It has been suggested that Dt is more effective in treatment of breast cancer than paclitaxel [
39] and in addition to its influence on microtubule assembly that culminates in a general cytotoxic response, Dt activity has been linked to increased activation of the apoptotic program and to changes of apoptotic marker expression [
40‐
43]. It may be these additional activities of Dt that combine with 267 to produce enhanced therapeutic effects.
It was important to demonstrate that the individual drugs within the 267/Dt combination exert benefits consistent with their individual mechanisms of action. For example, 267 activity can be linked to measured changes in P-AKT (ser473) levels and VEGF while Dt activity can be assessed by drug-mediated changes in cell architecture. ILK inhibition by 267 engenders dose dependent decreases in levels of P-AKT (Figures
1b,
4 and
5) and when 267 is added as a single agent it can inhibit VEGF secretion (Figure
6). Perhaps unexpectedly, single-agent 267 treatment also caused changes in cytoarchitecture and nuclear morphometry (Figure
7). This effect of 267 has not be reported previously, however, studies have provided evidence that ILK plays a role in cytoskeletal arrangement of actin through the regulation of proteins such as Rac and Cdc42 [
9,
44,
45]. Furthermore, siRNA mediated ILK silencing resulted in diminished cell spreading and actin cytoskeleton reorganization; results that help to explain ILK's role in the regulation of cancer cell motility and invasiveness [
46]. Recent evidence indicates a role for ILK in regulation of mitotic spindle organization [
47]. When this information is considered in light of the activity of Dt, one can speculate about the mechanism that may be promoting synergy when Dt is used in combination with 267. Studies have shown that cells treated with Dt exhibit a reorganization of the microfilament network [
36], disturbed microtubule structures, less F-actin stress fiber formation, decreased activation of Rac1/Cdc42, reduced cell motility, and an inhibition of angiogenesis [
48]. When considering the primary effect of Dt on the microtubule cytoskeleton of cancer cells, and based on the results summarized here it can be suggested that the combination of Dt and 267 may result in synergistic changes in tubulin, F-actin organization, and nuclear degeneration during apoptosis.
As indicated above, inhibition of ILK by 267 was expected to cause a decrease in P-AKT at serine 473. However, the effect of Dt on AKT has not been well studied. Studies have suggested that Dt can suppress the phosphorylation of AKT in lymphoma cell lines [
49] and lung carcinoma [
50]. Others have suggested that the AKT pathway can be activated by Dt [
51]. As shown in Figure
4, results obtained in several breast cancer cell lines indicate that Dt added at doses of up to 1 nM exerted no significant effect on P-AKT levels after an eight-hour exposure. Importantly, Dt potentiates the effect of 267 on P-AKT levels, at least in LCC6 and MCF-7 cell lines (Figures
4 and
5). Interestingly, this beneficial combination effect was not observed in the Her2 transfected variants of these cell lines, suggesting that phosphorylation of AKT does not play a role in the enhanced cytototoxicity seen when 267 is combined with Dt to treat the Her2 over-expressing cells.
It has also been established that one of the beneficial therapeutic effects of 267 is associated with its ability to inhibit VEGF secretion. More specifically, it has been reported that integrins cooperate with the VEGF receptors to promote angiogenesis in vascular endothelial cells [
52] and other studies indicate that ILK and PI3-kinase are involved in VEGF signaling pathways [
53]. Although not well studied, it has been suggested that Dt can influence vascularization
in vivo in a fashion that is related to VEGF signaling. More specifically, Murtagh and Schwartz [
54] have recently demonstrated that Dt can prevent VEGF-induced phosphorylation of focal adhesion kinase, Akt and endothelial nitric oxide synthase, effects that may be mediated by Dt mediated dissociation of Hsp90 from tubulin and subsequent Hsp90 degradation by ubiquination. Thus, it could be speculated that combinations of 267 and Dt would be of particular interest in the context of VEGF-induced tumor vascularization; where 267 would suppress VEGF production and Dt would mitigate signaling through any remaining VEGF. However, preliminary
in vitro studies summarized in Figure
6 suggest in the cell lines that express low levels of Her2 that the 267/Dt combination was less effective at inhibiting VEGF secretion then when 267 was used alone. Similar to the P-AKT (Ser473) results, when using VEGF secretion as an endpoint, the results obtained in the Her2 over-expressing cell lines differed from those obtained with cells that express low Her2 levels.
On the basis of VEGF secretion and P-AKT (Ser473) data we can conclude that the 267/Dt drug combination effects were dependent on Her2 expression. These differences encouraged us to assess the effect of 267 on Her2 signalling in the Her2-positive cell lines. Although not reported here, these studies demonstrated that 267 treatment induced a dose-dependent decrease in Her2 levels; an effect that could also be obtained when using siRNA to silence ILK. This unexpected effect of 267 on Her2-positive cell lines complicated the interpretation of results in these cells and for this reason the in vivo studies reported here focused on mice bearing orthotopically transplanted LCC6 cells, which do not express detectable levels of Her2.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JK designed and executed all experiments and data analysis and wrote the manuscript. CW contributed the LCC6Her2 cell line, executed confirmatory studies, and helped revise the manuscript. KF conducted initial exploratory studies to identify drug combinations. LE contributed to study design. TD contributed QLT0267. DW, and BS helped write the manuscript. KG gave clinical expertise. WD and SD helped revise the manuscript. MB is Chief Investigator, conceived the study design and helped write the manuscript.