Purpose: The high affinity receptor 1 (NTSR1) and its agonist, neurotensin (NTS), are correlated with tumor cell aggressiveness in most solid tumors. As chemoresistance and tumor aggressiveness are often related, we decided to study the role of the NTSR1 complex within platinum-based chemotherapy responses. In an ovarian model, we studied carboplatin because it is the main standard of care for ovarian cancer.
Experimental Design: Experimental tumors and in vitro studies were performed using SKOV3 and A2780 cells treated with carboplatin, with or without a very specific NTSR1 antagonist, SR48692. We measured the effects of these treatments on cell apoptosis and apoptosis-related proteins, platinum accumulation in the cell and nucleus, and the expression and localization of platinum transporters. NTS and NTSR1 labeling was measured in patients with ovarian cancer.
Results: SR48692 enhanced the response to carboplatin in ovarian cancer cells and experimental tumors. When SR48692 is combined with carboplatin, we noted a major improvement of platinum-induced DNA damage and cell death, as well as a decrease in tumor growth. The relationship of these results to clinical studies was made by the detection of NTS and NTSR1 in 72% and 74% of ovarian cancer, respectively. Furthermore, in a large series of high-grade ovarian cancer, NTSR1 mRNA was shown to correlate with higher stages and platinum resistance.
Conclusions: This study strongly suggests that the addition of NTSR1 inhibitor in combination with platinum salt–based therapy will improve the response to the drug. Clin Cancer Res; 23(21); 6516–28. ©2017 AACR.
Platinum salt–based therapy is used to treat almost 50% of cancer patients. It is proposed as an adjuvant, first-line, or in palliative therapy. Unfortunately, many patients relapse because of intrinsic, acquired resistance, or a weak sensitivity to the treatment. New therapeutic strategies to optimize the current treatment will add a real benefit to patient therapy. Neurotensin receptor 1 (NTSR1) is overexpressed in a large number of solid cancers, and contributes to tumor aggressiveness. We report that blocking NTS/NTSR1 complex improves the effect of carboplatin in ovarian cancer by enhancing the drug-to-target ratio (platinum binding to DNA), and consequently the toxic effect of the drug. A combined platinum salt–based therapy with NTS/NTSR1 inhibitor would be a suitable alternative to the current standard of care without adding toxicity to the treatment.
Ovarian cancer is the seventh most common cancer in the world and the eighth most frequent cause of cancer-related death among women (1). In the United States, 22,280 new cases and 14,270 deaths were estimated for 2016. This represents about 2.4% of all cancer-related deaths. Because this disease has no obvious specific symptoms, the majority of females with ovarian cancer are diagnosed at advanced stages, and the 5-year survival rate is 46.2% (2). Consequently, ovarian cancer is one of the most lethal gynecologic malignancies in women.
The standard treatment for advanced ovarian cancer patients is primarily debulking surgery and platinum-based chemotherapy combined with taxane (3, 4). The combination of carboplatin plus paclitaxel results in a complete response rate in 40% to 60% of the cases. However, more than 90% of these patients relapse after 2 years, and patients with recurrence become incurable due to the development of chemoresistance (5). Ovarian cancer remains a real clinical challenge, and despite the development of new therapeutic strategies, there is an urgent need to optimize the current treatment. To this regard, we suggest the use of neurotensin blockage to improve the sensitivity to platinum-based chemotherapy.
Neurotensin (NTS) is a 13-amino acid peptide acting as a neurotransmitter in the central nervous system and local hormone in gastrointestinal tract (6). NTS activates three subtypes of receptors. NTSR1 and NTSR2 belong to G-protein–coupled receptor family (7). NTSR3 (gp95/sortilin) is a sorting protein member of the family of Vps10p-domain receptor family (8). The complex of Neurotensin (NTS) and its high affinity receptor 1 (NTSR1) was shown to contribute to cancer progression (9). NTSR1 was found overexpressed in several types of solid cancers (10), in association with the dysregulation of the β-catenin pathway or epigenetic regulation (11–13). Moreover, the presence of NTS/NTSR1 complex enhanced the tumor growth and metastasis process in many solid cancers (7, 14–19). Our group demonstrated that this cell aggressiveness was due to the establishment of the EGFR autocrine activation, occurring under the sustained stimulation of NTSR1 (7, 12, 14).
Independently of the contribution of NTS/NTSR1 complex to cancer progression, it was recently shown that NTS plasma concentration is increased in rats treated with cisplatin and oxaliplatin for several weeks, causing specific damage of the sciatic nerve, and producing variable effects in motor and behavioral tests (20). We hypothesized that NTS/NTSR1 complex may contribute to chemotherapy resistance and that the blockage of NTS/NTSR1 may improve the response to platinum salt–based chemotherapy.
Until now, the most studied approach to antagonize NTSR1 activation is the use of the nonpeptidic component meclinertant (SR48692; ref. 21). This drug was developed to counteract the action of NTS on neurotransmitters in the brain (22). Meclinertant exhibits a nanomolar affinity for different tissues and cells from various species (21). This compound crosses the blood–brain barrier. It was reported to reverse most intracellular mediator signals linked to NTSR1 activation, such as intracellular Ca2+ mobilization, inositol monophosphate, cyclic GMP, and cyclic AMP activation (23). This component showed agonist effects on NTSR2, which may explain its partial or nonresponse on particular central nervous system effect induced by NTS as hypothermia or analgesia (24, 25). Meclinertant was also tested on nonselected patients with lung cancer, after a first line of chemotherapy using cisplatin combined with etoposide, but failed to demonstrate an increase in overall survival (NCT00290953). Globally, meclinertant is very efficient and specifically counteracts the physiologic action of NTS mediated by NTSR1, but lacks performance in pathologic situations as cancer where the NTS autocrine loop is very strong.
In this article, we showed that SR48692 enhanced the response to carboplatin in ovarian cancer cells expressing both NTS and NTSR1. SR48692 improved the effect of carboplatin by increasing the drug efficiency and reducing the survival action of NTS. We also showed the absence of NTS and NTSR1 expression in tissues from normal ovary samples, whereas these makers are present in approximately 70% of ovarian cancer.
Materials and Methods
Ovarian adenocarcinoma cells SKOV3 was purchased from ATCC and A2780 was purchased from European Collection of Authenticated Cell Cultures (ECACC). Cells were cultured in RPMI 1640 supplemented with 10% FBS (Gibco) and 2 mmol/L glutamine. SKOV3 was used at passages between 38 and 48 and A2780 was used at passages between 25 and 35. The corresponding NTSR1-overexpressing clone A2780-R1 was cultured in RPMI 1640 supplemented with 10% FBS, 2 mmol/L glutamine, and 0.5 mg/mL G418. All cells used in our research were mycoplasma-free as confirmed by the EZ-PCR Mycoplasma Test Kit (Biological Industries).
A2780 cells were transfected with NTSR1 expressing pcDNA3 vectors using Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol. Selection was performed with 1 mg/mL of G418 (Invitrogen). Stable transfectants were screened for NTS and NTSR1 expression by RT-PCR and immunofluorescence.
Cell viability assay
Tumor cell growth was evaluated by PrestoBlue Cell Viability Reagent (Invitrogen). Cell suspension (100 μL) containing 8,000 cells were seeded in 96-well plates, and the cells were treated accordingly. After 72 hours, 10-μL Presto blue was added to each well and the cells were incubated for 2 hours at 37°C. The fluorescence was measured by a multimode plate reader (EnSpire, PerkinElmer).
A2780 cells (2.5 × 106) were suspended in PBS and Matrigel (50%) (BD Biosciences) then inoculated in the right flank of nude mice (NMRI-Nude Foxn1). All the procedures were in accordance with the Guide of the Care and Use of laboratory Animals. Institutional Review Board approval was obtained by Le Comité d'Ethique en Expérimentation Animale Charles Darwin # B751201. Fourteen days after injection, four groups of six to seven mice were randomized as follows: 116 ± 10 mm3 for control group, 120 ± 11 mm3 for SR48692 group, 113 ± 10 mm3 for carboplatin group, and 110 ± 7 mm3 for carboplatin and SR48692 group. SR48692 was first dissolved in DMSO at the concentration of 10–2 mol/L. To avoid DMSO toxicity, SR48692 was further diluted in water (around 20×), strongly vortexed to form an emulsion. Mice were force-fed every day with this emulsion at the final dose of 1 mg/kg SR48692. In addition, at day 1, 3, 5, and 7, mice in the group carboplatin and carboplatin plus SR48692 were treated with carboplatin (4 mg/kg, i.p.), and the others were injected with PBS. Mice were also treated with a monoclonal NTS antibody at the dose of 10 mg/kg every 2 days. For details, see patent application to EPO under EP14305825.3.
RNA extraction, RT-PCR, and quantitative RT-PCR
Total RNA was extracted with guanidinium thiocyanate-phenol-chloroform acid method modified by Souaze and colleagues (26). For details, see primer sequence and quantitative RT-PCR in Supplementary Data.
Annexin V-FITC/PI dual staining analysis
SKOV3 and A2780-R1 cells were plated in 100-mm dishes in complete medium. After 24 hours, cells were treated with 5 μmol/L SR48692, 150 μmol/L carboplatin, or both for 72 hours for SKOV3 or 48 hours for A2780-R1. Cells were dissociated using Accutase solution (Sigma), washed with ice-cold PBS, resuspended in 100-μL incubation reagent (20 μL Annexin V-FITC and 20 μL PI diluted in 1-mL incubation buffer, Annexin-V-FLUOS Staining Kit, Roche), and then incubated for 15 minutes at room temperature in the dark. Samples were then analyzed by BD FACSCanto II (Becton Dickinson).
Hoechst 33258 staining assay
Cells were seeded on glass slides (12-mm diameter) in 24-well plate at the density of 3 × 104 cells/well in 500-μL medium and incubated 37°C with 5% CO2 overnight. Cells were treated for 48 hours then washed with PBS for three times and fixed with 4% paraformaldehyde for 20 minutes at room temperature. Cells were stained with Hoechst 33258 solution in the dark for 10 minutes at room temperature and imaged by a fluorescence microscope.
Caspase activity protease assay
Caspase activity was detected using ApoTarget Caspase Colorimetric Protease Assay Sampler Kit (KHZ1001, Invitrogen) and was performed according to the manufacturer's procedure (for details, see Supplementary Data).
Measurement of platinum uptake in whole cell and accumulation in DNA
Platinum accumulation in whole cell and DNA was determined by inductively coupled plasma mass spectrometry (ICP-MS) analysis. At each data point, cells were trypsinized, lysed in 1% Triton X-100, 0.1% SDS (w/v) buffer for 15 minutes on ice, and then incubated in 52.5% (v/v) nitric acid at 65°C for 2 hours. Finally, the acid concentration was reduced to 5% for storage. For DNA platinum accumulation, cells were digested overnight at room temperature in lysing buffer [50 mmol/L Tris-HCl pH = 8, 10 mmol/L EDTA, SDS 0.6% (w/v), 50 μg/mL RNase A, 400 mmol/L NaCl, 166 μg/mL proteinase K]. After digestion, the samples were homogenized in 1 mol/L NaCl by shaking. Samples were centrifuged at 3,000 rpm for 15 minutes at 4°C, and volumes of ethanol were added to the supernatants. The DNA was retrieved by winding the DNA around a glass pipette and then washed in 70% ethanol, air dried, and then resuspended in water. The results were normalized by the protein levels or DNA amounts.
Immunocytochemistry was performed according to the procedure previously described (7) and detailed in Supplementary Data.
SDS-PAGE and Western blotting analysis
Western blots were processed as described previously (7). Western blots were revealed with antibodies directed against anti-ATP7A (1:500), anti-ATP7B (1:500), anti-MRP2 (1:500), anti-BCL-2 (1:500), anti-BCL-xL (1:500), anti-BAK (1:1,000), and HRP-conjugated anti-actin purchased from Santa Cruz Biotechnology, and anti-active caspase-3 (1:1,000) from BD Pharmingen. Primary antibodies were incubated overnight at 4°C according to the manufacturer's instructions. Secondary anti-rabbit (Santa Cruz Biotechnology), anti-goat (Santa Cruz Biotechnology), or anti-mouse (Sigma) antibodies conjugated to HRP were used at 1:2,000 dilutions for 1 hour at room temperature and visualized by enhanced chemiluminescence (Pierce ECL 2 Western Blotting Substrate, Thermo Scientific).
Analysis of patient data
The mRNA expression was analyzed by (Agilent, Affymetrix HuEx, Affymetrix U133A. Ovarian serous cystadenocarcinoma TCGA dataset (http://cancergenome.nih.gov/) was obtained via cBioPortal (http://www.cbioportal.org; refs. 27, 28). Clinical data, main molecular features, NTSR1 and NTS mRNA z-scores (RNASeqV2), were available for 491 cases (29).
Population, clinical data, and tissues
Forty-six consecutive cases of ovarian tumor tissues with the main clinical data were retrospectively retrieved from the files of the Department of Pathology (CHU Cochin Port-Royal, AP-HP, Paris, France). Clinical and histologic characteristics of the patients are detailed in Supplementary Table S1. Ten nonmalignant ovarian samples were also retrieved (CHRU of Nancy, Nancy, France). The histologic subtype, the grade according to the criteria of the WHO classification, neoplastic emboli, local, and nodal invasion, were reviewed by two experienced pathologists (G. Gauchotte and P.-A. Just).
For all cases, histologic slides of primary tumors were obtained by paraffin wax–embedded tissues. Deparaffinized tissue sections (5 μm) were subjected to heat-induced epitope retrieval in citrate buffer (pH 6.0). The sections were labeled for the target proteins using the avidin–biotin–peroxidase complex method as described in ref. 30 (for details, see Supplementary Data). Cytoplasmic staining for NTSR1 and NTS were evaluated using a semiquantitative score. The staining intensity was graded as following: 1, weak; 2, medium; and 3, strong. A score was obtained by multiplying the percentage of positive cells by the intensity level (range, 0–300). Samples with score <50 were considered as negative or weak, and samples ≥50 moderate or strong.
Informed consent was obtained from all patients before surgery. The database including pathologic variables was established in accordance with the French data protection authority.
All statistical analyses were performed using GraphPad Prism (GraphPad Software, Inc.). Statistical analysis was carried out using t test or Bonferroni multiple comparisons test (****, P < 0.0001; ***, P < 0.001; **, P < 0.01; and *, P < 0.05). A P value of less than 0.05 was considered significant.
The comparison of NTS and NTSR1 mRNA z-scores or IHC semiquantitative score between clinicopathologic groups were performed using the Mann–Whitney U test (two groups) or Kruskal–Wallis test with Dunn postanalysis (more than two groups). A P value of 0.05 was accepted as significant.
NTSR1 antagonist enhances response to carboplatin treatment of the ovarian cancer experimental tumors
Expression of NTS and NTSR1 was evaluated by RT-PCR in two ovarian adenocarcinoma cell lines, SKOV3 and A2780. A strong expression of NTS was observed in A2780 cells, whereas a more moderate expression was detected for SKOV3 cells. A very weak expression of NTSR2 was detected in both SKOV3 and A2780 cell lines. In contrast, NTSR1 mRNA was only detected in SKOV3 cells (Fig. 1A). Tumors initiated with SKOV3 cells grow very slowly (31). For this reason, we chose to generate experimental tumors from A2780 cells, which exhibit palpable tumors within 10–14 days after cell injection.
As shown in Fig. 1B, the tumor growth rate of the control group was very high, and therefore the experiment was terminated after 15 days. In the carboplatin-treated group, the tumors were slightly responsive at the beginning of the treatment, but reached the same tumor volume as the control group at the end of the experiment. SR48692 treatment did not significantly affect the tumor growth, but when carboplatin was used in combination with SR48692, the tumor growth rate was significantly decreased. We confirmed that a decrease of NTSR1 activation improved the response to platinum salt therapy by using a mAb directed against NTS. Using the combination of cisplatin with NTS antibody, the tumor size is smaller than in the treatment cisplatin or NTS mAb alone (Fig. 1C).
Surprisingly, using the combination of carboplatin and SR48692, we observed a strong diminution in the size of the tumors generating the xenograft of A2780 cells, which do not express NTSR1 when cultured on plastic (Fig. 1A). However, when we analyzed the NTS and NTSR1 mRNA expression levels in the tumors, we observed an unexpected significant sixfold increase of NTSR1 mRNA levels as compared with those in cultured cells (Fig. 1D). In contrast, the levels of NTS mRNA remained similar between the tumor and cell cultured conditions (Fig. 1E). A more thorough analysis of mRNA levels from tumors in treated animals revealed that the NTS mRNA levels from tumors with carboplatin treatment were almost 10-fold higher than those from control mice (Fig. 1F). This result was confirmed at the protein level, as the labeling intensity for long fragment NTS was much stronger in tumors from mice treated with carboplatin (Fig. 1H). The labeling of NTS mature peptide is more diffuse but also more intense than in the control conditions. The combined treatment with carboplatin and SR48692 abolished this increase both at the RNA and protein levels (Fig. 1F and H). Finally, Ki67 labeling was less frequent and weaker in tumors treated with the combination as compared with control or carboplatin treatment alone.
NTSR1 antagonist improves the carboplatin response by inhibiting cell growth and survival
To elucidate the mechanisms of the enhanced effect of SR48692/carboplatin combination, in vitro studies were performed on either SKOV3 or A2780 cells ectopically expressing NTSR1. Indeed, as cultured A2780 cells do not express NTSR1, we established stable cell lines expressing NTSR1 and chose a clone with NTSR1 expression levels similar to SKOV3 cells, the A2780-R1 cells. The NTSR1 expression in A2780-R1 was confirmed by RT-PCR (Fig. 1A) and immunofluorescence (Supplementary Fig. S1A).
In SKOV3 cells, carboplatin inhibited cell growth in a dose-dependent manner, as measured with the Presto blue assay (Fig. 2A). The addition of SR48692 to carboplatin enhanced this effect, as the IC50 for carboplatin alone was 127.50 ± 13.75 μmol/L while the IC50 for the combination of carboplatin and SR48692 was 70.19 ± 6.12 μmol/L (Fig. 2B). We also performed cell viability assay on the A2780 wild-type cells and its corresponding NTSR1-overexpressing cells, A2780-R1. The wild-type cells showed a higher sensitivity to carboplatin than A2780-R1 cells, IC50 were 122.20 ± 13.23 μmol/L and 338.20 ± 61.61 μmol/L, respectively. SR48692 enhanced the response to carboplatin only in A2780-R1 cells to reach the IC50 value of 136.00 ± 30.60 μmol/L. In contrast, the sensitivity of A2780 wild-type cells to carboplatin was not increased significantly in the presence of SR48692 (Supplementary Fig. S1B).
Apoptotic cells were detected by Annexin V/PI dual labeling assay in the flow cytometry performed on SKOV3 and A2780-R1 cells. A higher number of cells were detected in Annexin V–positive quadrants (right up and bottom) when cells were treated with 5 μmol/L SR48692 and 150 μmol/L carboplatin (dose corresponding approximatively to the IC50 as determined above), as compared with the carboplatin treatment alone (Fig. 2C). The number of apoptotic cells was increased by twofold for the SKOV3 cells, and fourfold for the A2780-R1 cells when the combined treatment of carboplatin plus SR48692 treatment was compared with carboplatin alone treatment (Fig. 2D and E).
This finding was confirmed by examining nuclear morphologic changes in SKOV3 cells using Hoechst 33258 staining. Cells were scored as apoptotic if their nuclei presented chromatin condensation and nuclear beading or fragmentation. In cells treated with carboplatin for 48 hours, only several cells showed apoptosis-like changes. While cells treated with the combination of carboplatin and SR48692, more cells displayed morphologic changes as indicated by arrows (Fig. 2F). The apoptotic cells were counted and the apoptotic ratio was calculated in comparison with the total number of cells in each field. The percentage of apoptotic cells increased from 1.56% ± 0.10% for carboplatin-treated cells to 4.15% ± 0.25% for the combination treatment (Fig. 2G).
NTSR1 antagonist enhances carboplatin-induced cell death through caspase activation and modulation of Bcl-2 protein family expression
Carboplatin exerts its cytotoxicity via the formation of DNA lesions, which ultimately results in cell-cycle arrest and the induction of cell death through the apoptotic pathway. To elucidate the role of caspases in carboplatin-induced apoptosis in SKOV3 cells, we first examined the activation of executioner caspase-3 by Western blot analysis. As shown in Fig. 3A, when cells were treated with carboplatin, both the proenzyme of caspase-3 (32 kDa) and two active caspase-3 subunits (20 and 17 kDa) were detected after 48 hours (lane 3) of treatment. Furthermore, caspase-3 activation was more pronounced at 48-hour treatment (lane 4) when cells were treated with the combined treatment. Stimulation of the other members of the caspase family was measured by a caspase colorimetric protease assay kit. As shown in the Fig. 3B, all five caspases (2, 3, 6, 8, and 9) were activated by carboplatin, and in the presence of SR48692, the activation of these caspases was even significantly enhanced. We note that both initiator and effector caspases are overstimulated in the combined treatment. This regulation may contribute to the improvement of carboplatin response when combined with SR48692.
The fate of cells is determined by the balance between anti- and proapoptotic members of the Bcl-2 protein family. As expected, Bcl-2 and Bcl-xL, were downregulated, while the mRNA of Bak and BCL2L11 (encoding Bim, a BH3-only Bcl-2) were increased with the combined treatment as compared with carboplatin single treatment (Fig. 3C). The Bcl-2 family protein profile, Bcl-2 and Bcl-xL, followed a similar regulation as their respective transcript. An increase in Bak protein expression with the combined treatment was also observed (Fig. 3D).
NTSR1 antagonist facilitates nuclear platinum accumulation and increases the DNA double strand breaks
Platinum accumulation in whole-cell extracts and nuclei was compared in the presence of carboplatin or carboplatin plus SR48692 treatments. In SKOV3 and in A2780-R1 cells, the platinum content of the whole cell was increased by twofold after 6-hour exposure to the combined treatment as compared with carboplatin alone (Fig. 4A). No increase was observed in A2780 wild-type cells. To evaluate whether SR48692 facilitates drug access to the targeted DNA, platinum-DNA binding was assayed. After a 6-hour treatment, an increase of 2.02 ± 0.23 and 2.61 ± 0.33 fold of platinum-DNA binding was observed in SKOV3 and A2780-R1 cells, respectively (Fig. 4B). In A2780 wild-type cells, a slight but no significant increase of platinum content in the nucleus was seen when cells were treated with carboplatin in the presence of SR48692 (Fig. 4B).
It is documented that the phospho-SER-139 H2AX protein (γ-H2AX) is recruited at the site of DNA double strand breaks and thereby involved in DNA repair. Therefore, we analyzed its recruitment in SKOV3 cells after treatment with carboplatin alone or in combination with SR48692. The number, the size (as see in Fig. 4C, left), and the intensity of dots increased in the nucleus of cells treated with carboplatin and SR48692, or LF-NTS mAb as compared with carboplatin single treatment. Barely any labeling was detected in control SKOV3 cells, and cells treated with SR48692 or LF-NTS mAb alone. The intensity of γH2AX labeling was increased from 121.30% ± 5.83% for carboplatin-treated cells to 184.10% ± 8.41% for the SR48692 and carboplatin combined treatment and to 172.6% ± 20.65% for the LF-NTS mAb and carboplatin combined treatment as compared with the control cells (Fig. 4C, right).
NTSR1 antagonist reduces carboplatin efflux from the cells
The P-type transport ATPase family (ATP7A, ATP7B) and the multidrug resistance-associated protein 2 (MRP2) have been reported to be associated with the platinum efflux and consequently to platinum salt drug resistance (32). In SKOV3 cells, we found that the mRNA expression levels of ATP7A, ATP7B, and MRP2 were decreased when cells were treated with the combination of SR48692 and carboplatin (Fig. 5A). The decrease was even more pronounced at the protein level suggesting a decrease of the cellular efflux (Fig. 5B). ATP7A shuttles from the Golgi apparatus to the cell membrane when the copper concentration is increased in cells and contributes to the copper efflux (33). We investigated whether the localization of ATP7A was altered when SKOV3 cells were exposed to 500 μmol/L carboplatin for 1 hour. In the control cells and the cells treated with SR48692, ATP7A accumulated in the trans-Golgi network while the labeling was dispersed in the cytoplasm when cells were exposed to carboplatin (Fig. 5C). However, in the cells treated with both carboplatin and SR48692, the labeling of ATP7A was less dispersed in the cytoplasm as compared with cells treated with carboplatin alone (Fig. 5C). The quantification of the ATP7A immunolabeled particles was made according to their size (pixels) and showed that the number of particles was two- to threefold higher in cells treated with only carboplatin as compared with the cells treated with the combination of carboplatin and SR48692. This suggests that the mobilization of ATP7A in the cytoplasm, for the efflux of carboplatin, is antagonized by the NTSR1 antagonist.
NTS and NTSR1 are expressed in ovarian cancer
Labeling of NTS and NTSR1 was performed by IHC on 46 ovarian cancer specimens. Figure 6A shows a typical positive and negative labeling for both markers in low-grade serous carcinomas. The clinical and the histologic characteristics of patient are given in the Supplementary Table S1. The patients presented the conventional risk factors for ovarian adenocarcinoma: BRCA mutations, nulliparity, endometriosis history, menopause hormone therapy. The predominant pathologic subtype detected among the ovarian adenocarcinomas was serous differentiation (66.7%). The majority of the tumors were FIGO stage 3 (48.5%) or 4 (30.3%). This series is representative of patients with ovarian cancer.
In malignant tumors (borderline tumor and adenocarcinoma), the NTSR1 distribution was in most cases homogeneous (Fig. 6B). A Hirsch score superior to 50 was considered as moderately positively or strongly labeled. With this scoring, 72% and 74% of the cases were positively labeled with the NTSR1 and long-fragment NTS antibody, respectively. NTSR1 and NTS cytoplasmic localization was predominant and 67% of the samples were positive for both NTSR1 and NTS (Fig. 6B). The highest score for NTSR1 was found in the clear cell carcinomas. NTSR1 and NTS scores were also analyzed in nonmalignant tissues with a staining negative or very weak in nine of 10 samples (Fig. 6B).
NTSR1 expression was significantly higher in adenocarcinomas (102.1 ± 54.4) than in borderline tumors (67.7 ± 33.8) and normal ovaries (16 ± 23.5; P < 0.001). NTS expression was also significantly higher in adenocarcinomas (97.3 ± 48.2) than in borderline tumors (81.2 ± 55.6) and normal ovaries (28.5 ± 32.5; P < 0.001; Fig. 6C).
NTS and NTSR1 mRNA expression were analyzed in a cohort of 491 high-grade ovarian serous cystadenocarcinoma from the TCGA database. Figure 6D shows the distribution of the NTSR1 and NTS Z-score within the cohort. For both markers, this distribution was homogenously dispersed among the specimens. Kaplan–Meier analysis of the progression-free survival (PFS) for this dataset showed that a high NTSR1 mRNA expression (the z-score above the 90th percentile), was associated with a significantly worse prognosis (P = 0.019) with a median survival of 13.31 and 17. 51 months, respectively. FIGO stages were available in 484 patients. Twenty-four (4.9%) tumors were FIGO stage II (3 stage IIa, 4 stage IIb, 17 stage IIc), 381 (78.8%) tumors were FIGO stage III (7 stage IIIa, 21 stage IIIb, 353 stage IIIc), and 79 (16.3%) were stage IV. NTSR1 mRNA expression was significantly increased in higher stages (P = 0.01 for IV and IIIc stages vs. others; Fig. 2B; Supplementary Data). In the same vein, NTSR1 expression was higher within grade 3 as compared with grade 2 (Fig. 2C; Supplementary Data). Platinum status was available in 287 patients; 90 (26.1%) patients were resistant and 197 (57.1%) were sensitive. NTSR1 mRNA higher expression was significantly associated with platinum-resistant status (P = 0.0076), whereas NTS mRNA expression was not different for the resistant or the sensitive groups (Fig. 6E).
Interestingly, in this dataset, the expression of NTSR1 was inversely correlated with NTSR3 (P = 0.006; Fig. 2D; Supplementary Data). As a consequence, the expression of NTSR3, with a z-score above the 25th percentile was significantly correlated with a good prognosis of the progression-free survival (P = 0.0013) with a median survival of 14.72 and 11.24 months, respectively.
Surgery is the initial treatment for stage I–IVA ovarian cancer. It is usually followed by several cycles of chemotherapy using platinum-based drugs (mainly carboplatin) combined or not with other anticancer agents (paclitaxel or docetaxel). Nevertheless, most patients relapse because of a resistance or weak sensitivity to the platinum-based treatment. Therefore, improving the response to existing treatment by introducing a nontoxic sensitizing agent might optimize the current chemotherapy.
As the consequence of specific NTSR1 overexpression and activation in tumor cells, the NTS/NTSR1 complex enhances tumor progression by promoting proliferation (7, 14, 34, 35), metastasis (7, 14, 36, 37), as well as survival (38, 39). These observations translate in clinic to a worst prognosis in patients overexpressing NTSR1 (17, 36, 40). In this context, NTSR1-inhibiting agents are predisposed to improve the cancer outcome. Multiple studies showed that neurotensin receptor 1 antagonist, SR48692, inhibited proliferation of tumor cells in vitro as well as in vivo in small-cell lung cancer (19), pancreatic ductal carcinoma (41), breast cancer (42), colon cancer (18), and malignant melanoma (39).
In this study, we investigated the synergistic effect of a NTSR1 antagonist and platinum salt–based treatment in ovarian cancer cells and xenografted tumors in nude mice. We designed experimental tumors to mimic the clinical context and therefore waited until each tumor measurement reached 7 to 8 mm before starting treatment. Under these conditions, the tumor growth rates were three times less in the group treated with the combination carboplatin and SR48692 in comparison with the other groups. Clearly, blocking NTSR1 activation changed the cellular homeostasis of cancer cells to become more responsive to the chemotoxic drug.
Several possible mechanisms to boost the sensitization to chemotherapy were investigated. Carboplatin acts through the formation of DNA adduct, followed by apoptosis induction as a cellular response to DNA damage. Before reaching the DNA, carboplatin enters the cell via passive infusion or is facilitated by active uptake with a number of transport proteins. Platinum can be extruded from the cells by the GS-X pumps (MRP 1-5) after chelation with glutathione or via the copper efflux system (ATP7A/B; ref. 43). The decrease of cellular platinum accumulation is known to contribute to resistance to platinum-based drugs (44). In this view, overexpressing MRP2, ATP7A, and ATP7B were proposed to participate in the resistance mechanism (45–47). In particular, MRP2 (multidrug resistance-associated protein 2) showed a major role in chemoprotection by elimination of drug conjugates with glutathione (45, 48, 49). In our cellular model, SR48692 or carboplatin did not influence the expression of MRP2, ATP7A, or ATP7B, but the combination of NTSR1 antagonist with carboplatin strongly downregulates these three key players of drug efflux, and would explain the enhancement of carboplatin efficiency.
A major characteristic of NTS is to be a survival factor. This was demonstrated in several cancerous cellular models from diverse origins (9). In breast cancer cells, NTS agonist inhibited apoptosis induced by serum deprivation, and was accompanied by the enhancement of Bcl-2 gene transcription and protein cellular content (38). Platinum salt–based drug efficacy is obtained by the induction of apoptosis, often mediated concomitantly by the activation of several signal transduction pathways, including calcium signaling, death receptor signaling, and the activation of mitochondrial pathways. Cells escaping drug-induced apoptosis are closely associated with the development of platinum-based therapy resistance (50). The Bcl-2 family plays a crucial role in the regulation of apoptosis. It contains three subfamilies, the BH3 only family (Bid, Bim, Bad), the antiapoptotic family (Bcl-2, Bcl-xL), and the proapoptotic family (Bax, Bak). In our model, effector proteins in control of apoptosis are essentially unstimulated by carboplatin treatment alone, but in contrast, these effectors are strongly stimulated or downregulated accordingly in the combined treatment (in Fig. 3, see BCL-2, BCL-xl, BCL2L11, BAK, and BAX). As a consequence, the initiator (caspases 2, 8, 9) and executioner (caspases 3 and 6) caspases of the apoptotic pathway are also over stimulated when NTSR1 antagonist is combined with carboplatin. The significant caspase-2 activation suggests that it plays a critical role in the apoptosis in response to DNA damage caused by carboplatin and this response is enhanced by the addition of SR48692. We also noted the activation of caspase-8 pathway by the combined treatment, as compared with the carboplatin treatment alone, suggesting that the extrinsic apoptotic pathway was also implicated in better drug efficiency when associated with NTSR1 antagonist. This cascade merits a further exploration.
In summary, we have shown that a NTSR1 antagonist enhanced the response to platinum salt–based chemotherapy by improving the drug–target interaction and thereby enhancing the chemotherapy-induced apoptosis. These results suggest that patients with ovarian cancer expressing both NTS and NTSR1 could benefit from a combined treatment based on platinum salt drug associated with a blocking agent of the NTS/NTSR1 complex. Our initial studies suggest that this could concern the majority of the patients with ovarian cancer.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed by the authors.
Conception and design: A. Gompel, P. Forgez
Development of methodology: J. Liu, Z. Wu, G. Gauchotte
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J. Liu, M. Agopiantz, J. Poupon, Z. Wu, P.-A. Just, G. Gauchotte
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J. Liu, M. Agopiantz, G. Gauchotte, P. Forgez
Writing, review, and/or revision of the manuscript: M. Agopiantz, E. Segal-Bendirdjian, G. Gauchotte, A. Gompel, P. Forgez
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J. Liu, G. Gauchotte
Study supervision: A. Gompel, P. Forgez
We thank Dr. Neil Insdorf for his kind help in editing the manuscript. Many thanks go to Mr. J. Hotton for statistical analysis. We thank Sylvie Dumont and Fatiha Merabtene (UPMC Univ Paris 06, UMS_30 LUMIC, St Antoine Histomorphology Platform, F-75012) for their excellent assistance with the IHC. We thank Tatiana Ledent and other staff in the animal facility of St. Antoine hospital. We sincerely acknowledge the microscopy platform SCM (Service Commun de Microscopie - Faculté des Sciences Fondamentales et Biomédicales – Paris).
This work was supported by INSERM TRANSFERTPI-07563-A-09 (principal investigator: Forgez), SATT idf INNOV Paris France (principal investigator: P. Forgez), Fondation de France 2013 00038286 (principal investigator: E. Segal-Bendirdjian). China Scholarship council supported Dr. J. Liu's PhD study. SATT idf INNOV Paris France supported Z. Wu.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.