Background
Inflammatory breast cancer (IBC) is a very aggressive type of advanced breast cancer with a poor prognosis. IBC occurs typically in patients under the age of 50 and is often misdiagnosed as an infection since it does not present as a lump [
1]. The clinical symptoms of IBC involve the rapid onset of changes in the skin overlying the breast, including edema, redness, and swelling exhibiting a wrinkled, orange peel-like appearance of the skin known as peau d’orange [
2]. This peculiar presentation is associated with the invasion of aggregates of tumor cells, defined as tumor emboli, into the dermal lymphatics, where they obstruct the lymph channels [
3,
4]. Although IBC currently accounts for only 2–6 % of all breast cancer cases in the United States and up to 20 % of all breast cancers globally [
1,
5‐
7], its incidence is dramatically increasing [
2,
4]. Furthermore, due to its propensity to rapidly metastasize, IBC is responsible for a disproportionate number (15 %) of breast cancer-related deaths [
7‐
9]. IBC is either stage III or IV disease, depending on whether cancer cells have spread only to nearby lymph nodes or to other tissues as well. At the time of diagnosis, most IBC patients have lymph node metastases, and approximately 30 % have distant metastases in brain, bones, visceral organs, and soft tissue [
1]. Currently, there is no adequate adjuvant therapy to reduce the risk of recurrence and mortality in IBC patients.
We recently found that focal adhesion kinase 1 (FAK1 or PTK2) is amplified, upregulated and phosphorylated (active) in IBC [
10]. FAK1 is a nonreceptor tyrosine kinase that localizes to areas termed focal adhesions where the cell membrane attaches to the extracellular matrix. FAK1 activation relies upon autophosphorylation of the Tyr 397 site that is found in the N-terminal domain; FAK1 Tyr 397 binds various signaling proteins, including Src, PI-3 kinase, and Grb-7. FAK1 also binds epidermal growth factor receptor (EGFR), vascular epidermal growth factor receptor (VEGFR), p53, and other molecules that are critical for tumor growth and progression [
11]. It controls various cellular pathways, including proliferation, viability, and survival, and its overexpression has been linked to anoikis resistance [
12‐
15]. The activation and phosphorylation of FAK1 stimulated by many forms of oncogenic transformation provide a plausible mechanism for the anchorage-independent growth of cancer cells. FAK1 overexpression correlates negatively with patient outcome and is associated with increased cell migration, invasion, and metastasis; elevated FAK1 expression has been reported in multiple human epithelial tumors [
14,
16‐
20].
CEP-37440 is a potent ATP-competitive, highly kinase selective, and orally active inhibitor of FAK1 (enzymatic IC
50: 2.3 nM, cellular IC
50: 88 nM) and anaplastic lymphoma kinase (ALK) (enzymatic IC
50: 3.5 nM, cellular IC
50: 40 nM, cellular IC
50 in 75 % human plasma: 120 nM) [
21]. In addition to a favorable metabolic stability and pharmacokinetic profile preclinically, CEP-37440 is also a brain penetrant [
21]. In this study, we investigated the effect of CEP-37440 using human IBC cells and preclinical models of IBC. We found that CEP-37440 was able to inhibit the proliferation of certain IBC cells by decreasing the levels of phospho-FAK1 (Tyr 397); none of the cells expressed ALK. Studies using IBC xenograft models showed that CEP-37440 also effectively reduces the growth of the primary tumor xenografts and inhibits the development of brain metastases in mice.
Methods
Cell lines
The cell lines used are described in Table
1. The SUM149 and SUM190 cell lines were purchased from Asterand Inc. (Detroit, MI, USA). The MDA-IBC03 cells were obtained from W.A. Woodward, and KPL4 cells were obtained from N.T. Ueno, The University of Texas MD Anderson Cancer Center, TX, USA. All other cell lines, MDA-MB-231, MDA-MB-468, MCF-10A, and MCF-12A, were purchased from American Type Culture Collection (ATCC; Manassas, VA, USA). The new model of IBC, designated as FC-IBC02, was previously developed in our laboratory, through the isolation of tumor cells from the pleural effusion of an IBC patient, as described elsewhere [
10].
Table 1
Sensitivity of the cell lines to CEP-37440
FC-IBC02 | IBC | Neg | Neg | Neg | 91 | |
KPL4 | IBC | Neg | Neg | Post | ~900 | |
SUM190 | IBC | Neg | Neg | Post | 890 | |
MDA-IBC03 | IBC | Neg | Neg | Post | 1860 | |
SUM149 | IBC | Neg | Neg | Neg | >3000 | |
MCF-12A | Normal-like breast epithelial cells | Neg | Neg | Neg | 1516 | |
MCA-10A | Normal-like breast epithelial cells | Neg | Neg | Neg | 1700 | |
MDA-MB-231 | BC (non-IBC) | Neg | Neg | Neg | 1930 | |
MDA-MB-468 | BC (non-IBC) | Neg | Neg | Neg | 1275 | |
Reagents
CEP-37440 was synthesized and provided by Teva Branded Pharmaceutical Products R&D, West Chester, PA, USA. CEP-37440 has modest plasma protein binding, high intrinsic solubility, reduced lipophilicity, favorable microsomal metabolic stability across species, reduced capacity for drug–drug interaction, and possesses favorable oral bioavailability and a lower clearance rate in vivo across multiple species [
21]. For in vitro assays, CEP-37440 free base was dissolved in dimethyl sulfoxide (DMSO) at concentration of 4 mM. For in vivo studies, CEP-37440 tri-hydrochloride di-hydrate salt was used and the drug was dissolved in distilled water. The media for the different cell lines are described in (Additional file
12).
Total FAK1 and phospho-FAK1 (Tyr 397) assays
Cells were grown to 80 % confluence in 100 mm in diameter Petri dishes. Cells were collected in cold phosphate-buffered saline (PBS) by scraping from culture plates, centrifuged, and washed twice with cold PBS; proteins were extracted with 500 μl of cell extraction buffer (Invitrogen, Life Technologies, Carlsbad, CA, USA) containing a mixture of protease inhibitors (Protease Inhibitors Cocktail, Sigma-Aldrich, St. Louis, MO, USA) and 1 mM phenylmethanesulfonyl fluoride (PMSF). After extraction for 30 min on ice, the extracts were centrifuged at 14,000 × g for 10 min at 4 °C, and the supernatant was transferred to a fresh microcentrifuge tube and kept at −80 °C until use. The protein concentration in the cell extracts was determined using the BCA protein assay kit (Pierce, Rockford, IL, USA). Total FAK1 was studied in the cell protein extracts using FAK (Total) enzyme-linked immunosorbent assay (ELISA) kits (Invitrogen, Frederick, MD, USA); cell extracts were diluted 1:40 in dilution buffer as was recommended by the manufacturer; total FAK1 was expressed as ng total FAK1/μg protein. To study phospho-FAK1, FAK (pY397) ELISA kits (Invitrogen, Life Technologies, Carlsbad, CA, USA) were used and cell extracts were diluted 1:5 in dilution buffer; phospho-FAK1 (Y397) was expressed as units of phospho-FAK1/μg protein.
Total ALK and phospho-ALK (Tyr 1604) assays
Total ALK and phospho-ALK (Tyr 1604) were studied using PathScan Total ALK Sandwich ELISA and PathScan Phospho-ALK (Tyr 1604) Sandwich ELISA kits, respectively (Cell Signaling Technology, Inc., Danvers, MA, USA); protein extracts from Karpas 299 cells, a human T cell lymphoma cell line, were used as positive control.
Cell proliferation assays
For FC-IBC02, SUM190, MDA-IBC03, SUM149, MDA-MB-231, MDA-MB-468, MCF-10A, and MCF-12A, a total of 2000 cells were plated per well in a 96-well plate, except for KPL4 in which 1000 cells were plated. After the cells were plated, they were allowed to attach overnight and then treated with CEP-37440 during several time points (0 h, 24 h, 48 h, 72 h, 96 h, 120 h, 144 h, 168 h, and 192 h). Several CEP-37440 concentrations were evaluated (0, 3, 10, 30,100, 300, 1000, 2000, and 3000 nM), and each concentration was tested in quadruplicate; as controls, cells growing in regular media (without CEP-37440) and media with 0.075 % DMSO (vehicle) were included. After CEP-37440 treatments, the living cells were quantified using the CellTiter 96 Aqueous One Solution Cell Proliferation Assay kit (Promega, Madison, WI, USA) following the manufacturer’s instructions. Briefly, the media with the CEP-37440 was removed, and 100 μl of fresh media was added to each well; 20 μl 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) compound was added and the cells were incubated at 37 °C in a cell incubator for 4 h. The absorbance at 490 nm and 630 nm (reference wavelength) was measured with a 96-well microplate reader (iMark™, Bio-Rad Laboratories Inc., Hercules, CA, USA).
Gene expression profiling
FC-IBC02 cells were treated with 1000 nM CEP-37440 for 48 h and RNA was isolated. Affymetrix Genechip® Human Gene 1.0 ST array (Affymetrix Inc., Santa Clara, CA, USA) were used for gene expression studies (Additional file
12). Data analyses were performed using GeneSpring software 13.1 (Agilent Technologies, Inc., Santa Clara, CA, USA). The criteria for differentially expressed genes were set at ≥ 2.0-fold changes. Statistical analysis was performed to compare two groups using
t test unpaired with a
p value less than or equal to 0.05. A heat map was generated from the differentially expressed gene list. The list of differentially expressed genes was loaded into Ingenuity Pathway Analysis (IPA) 8.0 software (
http://www.ingenuity.com) to perform biological network and functional analyses.
In vivo studies using SCID mice
Studies were approved by the Institutional Animal Care Committee at Thomas Jefferson University. A total of 10
6 cells were suspended in 100 μl PBS, mixed with 100 μl Matrigel (BD Biosciences, Bedford, MA, USA), and injected into the fourth left inguinal mammary fat pad of severe combined immune-deficient (SCID) mice. The animals were palpated daily for detection of tumor development and, once the breast tumor xenografts reached approximately 50–100 mm
3 (approximately 20–30 days postinjection), the mice were randomly allocated into groups. Two doses of CEP-37440 were tested for mice harboring FC-IBC02 or SUM149 breast tumor xenografts; the mice allocated to treatment received either 30 mg/kg twice a day (bid) or 55 mg/kg bid by oral gavage in a volume of 100 μl, 5 days/week for 35–40 days. For mice harboring SUM190 breast tumor xenografts, only the higher CEP-37440 dose (55 mg/kg bid) was tested. The CEP-37440 doses were chosen based on preliminary experiments in order to achieve an optimal plasma concentration–response relationship [
21]. Breast tumors were measured using a vernier caliper, and tumor volumes were calculated using the following equation: V = [(L1 + L2)/2] × L1 × L2 × 0.526 where L1 and L2 are the length and width of the tumor. After 40 days of treatment or when the primary tumor reached a volume of approximately 1 cm
3, the animals were euthanized by carbon dioxide (CO
2) inhalation. Breast tumors and other organs (lungs, heart, liver, spleen, brain, ovaries, kidneys, and lymph nodes) were removed, fixed in 10 % neutral-buffered formalin and paraffin-embedded for histological examination (Additional file
12).
Statistical analyses
For the analyses of the cell proliferation data, the log-transformed response measures (Abs 490 nm and Abs 630 nm) were modeled using the linear mixed effects (LME) model adjusting for correlations between repeated measures over time. The fixed effects included the ten concentrations and linear time trends. For the analyses of in vivo tumor growth data, the log-transformed tumor volumes were modeled using LME models adjusting for correlations between repeated measures from the same animal. The fixed effects included the control group and treatment groups (30 mg/kg CEP-37440, and 55 mg/kg CEP-37440), and linear and quadratic time trends. The LME models included either only linear terms or both linear and quadratic terms as appropriate for specific time-dependent trends. Percent tumor growth inhibition (% TGI) was calculated as follows: 100 × [(tumor volume of the control group at the end of treatment − tumor volume of the treated group at the end of the treatment)/tumor volume of the control group at the end of the treatment] = 100 × (1-exp (mean difference in log volumes at the end of the treatment), where the mean difference in log volumes at the end of the treatment was estimated from the fitted LME models. The data were analyzed using R package ‘nlme’ (The R Foundation for Statistical Computing
http://www.R-project.org).
Discussion
Our studies showed that low concentrations of CEP-37440 specifically decreased the proliferation of the IBC cell lines FC-IBC02, SUM190, and KPL4 without affecting proliferation of normal breast epithelial cells. These IBC cells were sensitive to low concentrations of CEP-37440 and their sensitivity was independent of the cell subtype (triple-negative or ErbB2-positive). At higher concentrations, CEP-37440 also inhibited the proliferation of the IBC cell line MDA-IBC03 and the triple-negative non-IBC cell lines MDA-MB-231 and MDA-MB-468. Since CEP-37440 is a dual inhibitor of FAK1 and ALK, we studied the expression of both proteins in the IBC cell lines and found that they expressed FAK1 but not ALK. We demonstrated that CEP-37440 decreased the cell proliferation of FC-IBC02, SUM190, and KPL4 by blocking the autophosphorylation kinase activity of FAK1 (Tyr 397).
We previously showed that IBC cells exhibit amplification of the chromosomal arm 8q where FAK1 is located (8q24.3) [
10]. However, gene expression studies using arrays did not show any correlation between copy number and RNA expression in these IBC cell lines [
10]. Furthermore, our present work did not find a correlation between elevated total FAK1 expression at the protein level and FAK1 gene amplification in either IBC cells or triple-negative non-IBC cell lines. Increased dosage of the FAK1 gene is invariantly observed in the cell lines derived from human cancers of lung, breast, colon, and invasive squamous cell carcinomas [
22]. However, elevated FAK1 protein expression is not always correlated with amplification of the FAK1 gene. In human head and neck squamous cell carcinoma, not all cases with an amplification of the FAK1 gene display FAK1 protein overexpression, implicating a sophisticated posttranscriptional regulation involved in FAK1 expression and function [
23].
In vivo studies showed that CEP-37440 significantly decreased breast tumor growth in the SUM190 and FC-IBC02 mouse xenograft models, and the tumor growth inhibition was dose and time dependent. Mice harboring SUM190 and FC-IBC02 breast tumor xenografts showed 79.7 % and 33 % TGI, respectively, when treated with 55 mg/kg bid CEP-37440 for 35–40 days. The tumor breast xenografts of mice treated with CEP-37440 showed lower levels of phospho-FAK1 (Tyr 397) than the breast tumor xenografts in the control groups (data not shown). Although the IBC cell line SUM149 showed in vitro only a slight response to high concentrations of the drug, CEP-37440 reduced the size of the primary tumor in the mice harboring SUM149 breast tumor xenografts, and results were better with 30 mg/kg bid than 55 mg/kg bid CEP-37440. Since there was no total regression of the primary tumor or metastatic sites in the lungs in mice harboring IBC xenografts, our results suggest that a combination therapy approach would be more effective for IBC patients than CEP-37440 alone. Preclinical studies showed that the combination therapy approach using inhibitors of FAK1 with other signaling pathways increased the efficacy of single inhibitors [
22,
23]. These types of combinatorial studies have not been conducted with CEP-37440 in preclinical models of IBC or triple-negative breast cancer (TNBC).
None of the FC-IBC02 breast xenograft mice treated with CEP-37440 developed brain metastases in contrast to 20 % in the control group. This data suggests that CEP-37440 is able to cross the blood-brain barrier. Related work by another group found that TAE226, another small molecule inhibitor targeting the ATP-binding site of FAK1, increased apoptosis of glioblastoma, an infiltrative brain tumor, and inhibited tumor growth [
24].
The overexpression and phosphorylation of FAK1 on Tyr 397 is frequently associated with tumor metastasis as well as poor patient prognosis [
25‐
28], indicating a critical role for activated FAK1 in tumor progression and malignancy. Tyrosine 397 is the main autophosphorylation site of FAK1, leading to activation of its intrinsic kinase function as well as its downstream signaling players, and providing a high-affinity binding site for the SH2 domain of Src family kinases [
29,
30]. Furthermore, phospho-FAK1 binds to the p85 subunit of phosphatidylinositol 3-kinase (PI3K) at the Tyr 397 autophosphorylation site. Phospholipid production stimulated by phospho-FAK1 and activation of PIK3 can stimulate AKT kinase, which inhibits apoptosis by regulating various cell death cascade proteins [
31]. In addition to its roles as a cytoplasmic kinase, recent studies revealed that FAK1 can translocate to the nucleus where it can influence the expression of chemokines, which are secreted to the surrounding environment [
32]. These proteins recruit regulatory T cells (Treg) into tumors, which in turn suppress the CD8 antitumor response. Therefore, nuclear FAK1 signaling in cancer cells can help establish an environment within the tumor that supports survival and growth. FAK1 kinase inhibitors target mechanisms of immune suppression and may therefore represent a form of effective “immune-modulatory” therapy that reduces regulatory T cells in the tumor environment [
32]. We showed that CEP-37440 treatment of the IBC cell line FC-IBC02 increased the expression of multiple genes related to interferon signaling and cytokines.
In a recent study, FAK1 expression was analyzed in breast primary tumor samples from stages III–IV patients and a significant association between high FAK1 expression in the primary tumor, lymphovascular invasion, and triple-negative phenotype was found. In addition, a strong positive correlation was observed between high FAK1 expression and shorter overall survival and progression-free survival in patients with metastatic tumors [
33]; however, it was not specified by the authors if IBC samples were included in the study. The role of ALK in IBC is controversial, with some studies showing amplification of ALK in IBC and in others showing no amplification [
34,
35], and the present work showed that none of the IBC cell lines expressed the ALK protein.
FAK1 is one of the most attractive tyrosine kinase targets in cancer therapy since it plays a role in signal transduction and in the development of numerous human tumors, including breast, colon, thyroid, prostate, pancreas, and brain cancers [
36]. Besides CEP-37440, several other small molecule inhibitors of FAK1 that target the ATP-binding site and block FAK1 kinase activity have been developed. One of these inhibitors, PF-562,271 from Pfizer decreased tumor growth in multiple xenograft models [
37,
38], and was shown to be effective in phase I trials [
39]; however, it showed nonlinear pharmacokinetics and was henceforth discontinued. VS-6063 (defactinib) from Verastem showed more favorable pharmacokinetics, and in a phase I trial it demonstrated a favorable response in some patients with ovarian, colorectal, or bile duct cancer. Other FAK1 inhibitors, including VS-4718 (or PND-1186) [
40], VS-5095, and GSK2256098, are in early clinical trials.
Finally, our in vivo experiments demonstrated that animals treated with CEP-37440 did not experience reduced body weight or hair loss, suggesting that CEP-37440 was not toxic for mice at a dose of 30 or 55 mg/kg/bid. CEP-37440 is currently in a phase I dose-escalation clinical trial in patients with solid tumors (
www.clinicaltrials.gov NCT01922752). Based upon the preclinical studies described above and the data available for other FAK1 inhibitors, CEP-37440 may have clinical applications as part of a combinatorial therapy against IBC.
Competing interests
SVF received funds from Teva Pharmaceuticals to perform part of these studies. GRO and BAR were employees of Teva Pharmaceuticals. All remaining authors have declared no competing interests.
Authors’ contributions
IS, MA, and LDA carried out the in vitro assays and animal studies. IC performed statistical analysis of the data and helped in the interpretation of the results. SA performed the microarrays and analysis of microarray data. GRO and BAR participated in the design of the study and interpretation of the results. MC participated in the design and interpretation of the results. SVF coordinated the study, contributed to acquisition of data and interpretation of the results. The manuscript was written by BAR and SVF. All the authors were involved in revising the manuscript critically for important intellectual content. All authors have read and approved the final manuscript.