Background
Ovarian cancer is the leading cause of death from gynecological cancers and the patients are commonly diagnosed late with advanced disease. In general, the patients respond well to the primary treatment involving cytoreductive surgery and chemotherapy. However, more than 70% of the patients relapse, and in the recurrent disease, resistance to chemotherapeutic drugs is common[
1,
2]
. New targeted therapies are under evaluation, and immunotoxins (ITs) may represent an interesting alternative. ITs consist of an antibody, that with high affinity binds to the target antigen on the cancer cell surface, and a covalently bound toxin. Our MOC31PE immunotoxin binds to the cell surface antigen EpCAM, which is expressed by the majority of epithelial cancers including ovarian carcinomas. Upon internalisation
Pseudomonas exotoxin A (PE) inhibits protein synthesis by ADP-ribosylation of elongation factor 2 and induces apoptosis. EpCAM is a transmembrane glycoprotein, functioning as an epithelial-specific cell-cell adhesion molecule and may be involved in cellular signaling, migration, proliferation, and differentiation[
3]. Recently, it has been suggested that EpCAM is a cancer stem cell marker and may be expressed by cells undergoing epithelial to mesenchymal transition (EMT), lacking other epithelial markers[
4]. EMT-like cellular processes may be important during cancer metastasis, and EpCAM is thus an excellent candidate for therapeutic targeting of epithelial cancers. In a retrospective study of 500 ovarian cancer patients, EpCAM showed consistently high expression across different tumor stages and subtypes[
5] and the protein was over-expressed in cancerous tissues compared with non-cancerous ovarian surface epithelium and inclusion cysts[
6]. Notably, MOC31PE also induces cell death in chemotherapy-resistant cancer cells[
7] and may hence be used in patients with recurrent disease lacking other therapeutic options.
The immune suppressor cyclosporin A (CsA) was introduced in combination with IT to inhibit the host immune response during repeated IT administrations. In parallel with reduced anti-IT antibody production, synergistic cytotoxic effects were observed
in vitro and
in vivo[
8]. The immunosuppressive effect of CsA is caused by binding to cyclophilin A (CypA)[
9]. This complex binds and inhibits calcineurin a key enzyme for IL-2 production in T-cells. CypA over-expression has been reported in many human cancers and has also been suggested as a potential therapeutic target[
10]. Interestingly, CsA has been reported to reverse chemotherapeutic resistance in patients with recurrent ovarian cancer[
11,
12]. In the present work, we have studied the effects of MOC31PE treatment alone and in combination with CsA on protein synthesis, cell proliferation, viability, and migration on the ovarian cancer cell lines B76[
13] and HOC7[
14], which both express high amounts of EpCAM. Furthermore, MOC31PE-induced alterations in gene transcription were quantified in two different PCR-arrays: Cancer Pathway Finder and Tumor Metastasis.
Materials and methods
Materials
RPMI-1640, PBS, Glutamax, and Hepes were obtained from Lonza (Austria). Fetal calf serum was purchased from PAA (GE Healthcare, UK), MEM w/o leucine, 0.25% Trypsin/EDTA from Gibco, and YoYo-1 fluorescent dsDNA staining from Molecular Probes (Life Technologies, UK), and tritiated Leucine from Perkin Elmer (Waltham, MA). Cyclosporine A was purchased from Calbiochem (San Diego, CA) and dissolved in ethanol to 8.3 mM stock solution. The GenElute Mammalian total RNA kit and general laboratory chemicals were from Sigma Aldrich (St. Louis, MO), the Cell Titer 96 AqueousOne solution (MTS) cell proliferation assay was from Promega (Madison, WI). RT2 Profiler PCR Array System, including the cDNA synthesis kit, and SYBR green were from SABiosciences (Qiagen Nordic). Chemicals for validation of gene expression were from Applied (Life Technologies, UK). Plastic ware for cell culture was from Nunc (Thermo Scientific), gels and buffers for protein electrophoresis from Life Technologies, HRP-conjugated antibodies from Dako (DK), and chemiluminescent super-signal substrate from Thermo Scientific.
Cells and immunotoxin
The human ovarian cancer cell lines B76[
13] and HOC-7[
14] were a gift from Dr C. Marth (Innsbruck Medical University, Innsbruck, Austria). In this study B76 was our main cell line and HOC-7 was used to confirm key results. The cell lines were cultivated in RPMI 1640 medium added Glutamax, Hepes and 8% heat-inactivated fetal calf serum. The monoclonal antibody MOC31[
15] binds epithelial cell adhesion molecule (EpCAM, CD326) and was conjugated to whole
Pseudomonas exotoxin A as previously described[
16].
Protein synthesis and cell viability
The [
3H]-leucine incorporation assay was used to quantify protein synthesis[
16] and the Cell Titer 96 AqueousOne solution (MTS) assay was used to determine cell viability as previously described[
17].
Cell proliferation, membrane damage and scratch-wound healing in the IncuCyte
Cells were seeded in 96 well plates and grown to ≈ 50% confluency, transferred to the IncuCyte (Essen BioSciences, Ann Arbor, Mi) after the medium was replaced with fresh medium with or without IT and/or CsA. Membrane damage was measured after adding YoYo-1, a dye that emit fluorescence when it binds to double-stranded DNA. The cytotoxic index is defined as the number of fluorescent objects in a well, divided by the total number of fluorescent objects obtained after 0.1% Triton X-100 is added to open all cells in the well. For migration studies, the wound maker tool was used to make scratch wounds in confluent cell culture monolayers in 96 well image-lock plates (Essen BioSciences). Plates were incubated in the IncuCyte for 24 h and an integrated metric called relative wound density (RWD) was used to quantify effects on migration. This metric measures the cell density in the wound area relative to the cell density outside the wound area.
RNA isolation and PCR array analyses
The cells were seeded in 6 well plates, grown to ≈ 80% confluency and treated for 24 h before RNA was isolated from adherent cells using the GenElute Mammalian total RNA kit (Sigma Aldrich) and quantified in a Picodrop spectrophotometer (Picodrop Ltd, UK). RNA isolated for PCR array assays was treated with DNase I (Invitrogen) and the RNA quality was checked in the UV spectrophotometer. For cDNA synthesis (1 μg/reaction) the RT
2 first strand kit from SABiosciences was used. The resulting cDNA was diluted and qPCR was run as described in the PCR array protocol (SABiosciences RT
2 Profiler PCR Array System) using a BioRad ICycler. Gene expression was tested using either Cancer Pathway Finder (untreated, IT 10 ng/ml) - or Tumor Metastasis (2 μM CsA, CsA + IT 10 ng/ml) array. There are primers for 84 test genes and 5 reference genes (B2M, HPRT1, RPL13A, GAPDH, and ACTB) on each 96-well plate. Data analysis was performed as described in the protocol from the manufacturer and by using their PCR Array Data Analysis Web portal (
http://www.SABiosciences.com).
Validation of PCR array data
Gene expression was validated in independent experiments with RNA isolated as described above. The high capacity RNA to DNA master mix was used for cDNA synthesis (1 μg RNA/ reaction). Gene expression was measured using qPCR analyses with TaqMan probes using the 7500 Real Time PCR machine (Applied Biosystems). Each sample was tested in duplicate. Fold change in expression was calculated using the comparative C
t method with RPL37A as a reference gene since the expression of this gene was similar in control and experimental groups. The gene list and corresponding probes are shown in Additional file
1: Table S1.
Subcellular fractionation, gel electrophoresis, and antigen detection
Cells were grown to 70-80% confluency in 75 cm
2 flasks and treated with MOC31PE and/or CsA for 24 h. The cells were washed with cold PBS and lysed in 500 μl SF buffer (250 mM sucrose, 20 mM Hepes, 10 mM KCl, 1.5 mM MgCl
2, 1 mM EDTA and 1 mM EGTA, pH 7.5) and the protease inhibitor cocktail was added (MiniComplete, Roche). Cells were scraped from the flasks and the lysates were passed through 25G needles 10 times, and incubated on ice for 20 min. The nuclear pellet was centrifuged out at 720 g for 5 min and the resulting supernatant centrifuged at 10000 g for 10 min to separate the cytosolic (supernatant) and mitochondrial (pellet) fractions. Pellets were washed with 500 μl SF buffer, passed through 25G needles 10 times and re-centrifuged. Finally, the pellets were resuspended in 50 μl lysis buffer (10 mM Tris pH 7.5, 1% SDS, 1 mM Na
3VO
4, 0.1% Triton X-100, and 10% glycerol) and briefly sonicated. For total cell lysates, cells were lysed in boiling lysis buffer as previously described[
17]. Proteins were resolved on 4-12% Nu-PAGE gels and blotted onto PVDF membranes for antigen detection. The purity of the fractions was validated with antibodies detecting α-tubulin (cytosol, Cell Signaling), lamin B1 (nucleus, Abcam), and F
1F
0-ATP synthase (mitochondria, Calbiochem). NR4A3 in the fractions was detected on separate blots using a polyclonal anti-NR4A3/NOR-1 antibody (Novus Biologicals). Chemiluminescence signals were recorded using the G:Box system with a CCD camera from SynGene and quantified using the provided GeneTool software.
Statistical analyses
Statistical significance was evaluated with two-tailed Students t-test except for qPCR validations where non-parametric Mann–Whitney tests were used. In both tests p-values at 0.05 were considered statistically significant.
Discussion
The major limitation to curative therapy for ovarian cancer is acquired drug resistance to the chemotherapeutic agents used, such as Carboplatin and Paclitaxel. An additional drawback is the induced severe side-effects, mainly caused by the non-cancer cell specificity of the agents, reducing the patients’ quality of life. It is therefore necessary to identify novel drugs, which circumvent these disadvantages for successful treatment of ovarian cancer. In the present study, we have demonstrated in several different assays that the MOC31PE effectively inhibits protein synthesis, proliferation and cell survival of ovarian cancer cells, B76 and HOC7. Previously, we have reported in other tumor types synergistic cytotoxic effects of combining MOC31PE and CsA
in vitro and in an experimental metastasis model in animals[
8]. In agreement with previous results in other tumor types, these effects are potentiated when cells are simultaneously exposed to the immunosuppressant CsA.
The MOC31PE only binds to and kill cells expressing the antigen EpCAM, which is expressed in more than 90% of all epithelial ovarian carcinomas and to a negligible amount on normal cells, reducing the possibility of IT induced side effects in patients. In a recently conducted Phase I clinical study with MOC31PE, the tolerable profile was satisfactory (Andersson et al., in preparation), which is encouraging for clinical evaluation of MOC31PE against ovarian carcinoma. Interestingly, Phase I and II trials with CsA have shown beneficial effects on chemoresistance in patients with ovarian cancer[
11,
20] indicating that the combination of MOC31PE and CsA could be used for recurrent ovarian cancer.
Gene expression analysis was performed to identify affected signaling pathways induced by the treatments and several interesting candidate genes were found. In the Cancer Pathway Finder array, the majority of the genes affected by MOC31PE were related to angiogenesis, reflecting the importance of this cancer pathway in B76 cell growth. The two genes with the highest fold increase in expression on the array, PDGFβ and THBS1, was confirmed by qPCR. The PDGF network was recently identified as a biomarker for prognosis in ovarian cancer where higher levels of PDGF pathway activity were associated with reduced survival[
21]. The angiogenesis inhibitor Bevacizumab (Avastin), that binds to VEGF A, is an used molecular target agent in ovarian cancer[
22]. Given the importance of the PDGF pathway, targeting of VEGF, PDGF, and FGF at the same time may be more effective than targeting only VEGF[
23]. THBS1 was the first endogenous angiogenesis inhibitor identified[
24]. A role in cancer progression and cancer inhibition has been ascribed to the protein, and different effects of THBS1 depending on the phase of the progression have been suggested[
25]. In an early stage, high stromal expression of THBS1 inhibits tumor growth whereas later in the vascularized tumor THBS1 may increase the adhesive properties of tumor cells or modulate extracellular matrix proteins thereby promoting tumor invasion. We observed that CsA mono-treatment inhibited migration and reduced expression of some transcripts, including THBS1 in addition to potentiating IT effects. Calcineurin, the phosphatase inhibited by CsA, has been reported to regulate transcription of CTSK[
26] and CXCR4[
27]; two of five other affected genes. The inhibition of B76 cell migration by IT + CsA treatment may be a result of reduced THBS1 and/or MMP9 protein levels since increased transcription cannot be accompanied by increased translation due to IT-induced protein synthesis inhibition. In the tumor metastasis array mainly increased gene expression was seen when comparing CsA alone versus CsA + MOC31PE treatment of B76 cells. Examples of genes influenced are the metastasis suppressor KISS1 and its receptor. In ovarian carcinoma the increased expression of KISS1 has been shown to inhibit cell migration[
28]. This might support the results from the scratch-wound healing assay showing decreased migration in the B76 cells treated with MOC31PE alone or MOC31PE + CsA. Higher expression of KISS1 may also sensitize cancer cells for chemotherapy[
29]. Thus our results might support a contribution of MOC31PE as a supplement also to reduce chemotherapy resistance in ovarian cancer treatment.
The largest up-regulation was observed for the nuclear hormone receptor NR4A3, a member of the NR4A subfamily with poorly understood biological function and unknown physiological ligands[
30]. Depending on the context, NR4A transcription factors may be pro-survival factors or induce cell death[
31]. Knock-out mice without NR4A3 (Nor-1) and NR4A1 (Nur77) developed spontaneous acute myeloid leukemia[
32] suggesting tumor suppressing effects. In cancer cells, growth factors and mitogens induce expression of these transcription factors suggesting a role in cancer growth[
31]. However, induction of NR4A1 also occurs in response to apoptosis inducing factors in cancer cells. When translocated to mitochondria NR4A1 binds BCL-2, thereby inducing apoptotic cell death[
31] and during apoptosis in thymocytes mitochondrial targeting of NR4A3 was observed[
33]. In B76 cells, the majority of the NR4A3 protein was located in the cytosol. Two main changes in intracellular distribution were observed. MOC31PE or CsA shifted the protein to the mitochondrial fraction compatible with induction of apoptosis. Especially in MOC31PE + CsA treated cells increased NR4A3 was detected in the nuclear fraction. Increased amount of 60 kDa protein points to increased transcription of its target genes. Since increased 55 kDa protein in the nuclear fraction was accompanied by increased mitochondrial marker protein, and the nuclear fraction was pelleted at low speed, this implies that the mitochondrial mass has increased or that mitochondria have fused to larger structures. This is most likely an effect of the ongoing cell death. The increase in NR4A3 transcript, signals a need for NR4A3 protein synthesis. No corresponding increased NR4A3 protein was detected as IT inhibits protein synthesis, but translocation of NR4A3 to mitochondria enriched fractions suggests a role for this protein in MOC31PE-induced cell-death.
In summary, these results show that a PE-containing IT, MOC31PE, induces transcription of mRNAs for genes involved in angiogenesis and tumor metastasis. In addition, the therapeutic use of MOC31PE alone or in combination with CsA may provide an approach to the treatment of recurrent/chemoresistant ovarian carcinoma, but further investigation is needed to elucidate the effect of MOC31PE and CsA in ovarian cancer models in vivo.
Competing interests
The authors declare that they have no competing financial or non-financial competing interests.
Authors’ contributions
MTW provided, analyzed, and interpreted experimental data and drafted the manuscript. HB was involved in experimental data acquisition, ØF and KF participated in interpretation of the data and revisions of the manuscript, YA substantially contributed to the conception and design of the study, interpretation of experimental data and major revisions of the manuscript. All authors have approved the final version of the manuscript.