Advertisement

Functional Analysis of the Adrenomedullin Pathway in Malignant Pleural Mesothelioma

      Abstract

      Introduction

      Malignant pleural mesothelioma (MPM) grows aggressively within the thoracic cavity and has a very low cure rate, thus highlighting the need for identification of new therapeutic targets. Adrenomedullin (AM) is a multifunctional peptide that is highly expressed in several tumors and plays an important role in angiogenesis and tumor growth after binding to its receptors, calcitonin receptor–like receptor/receptor activity–modifying protein 2 (CLR/RAMP2) and calcitonin receptor–like receptor/receptor activity–modifying protein 3 (CLR/RAMP3).

      Methods

      Real time quantitative reverse transcriptase polymerase chain reaction (RT-PCR) was used to assess the steady-state levels of AM, CLR, RAMP2 and RAMP3 messenger RNA (mRNA) transcripts in normal pleural tissue (n=5) and MPM (n=24). The expression of these candidates at protein level was revealed by immunohistochemistry. We also characterized the expression and regulation by hypoxia of AM system in MPM cell lines and MeT-5A cells. In vitro and in vivo studies were performed to determine the functional role of AM system in MPM.

      Results

      In this study, real-time quantitative reverse transcriptase polymerase chain reaction showed twofold to 10-fold higher levels of AM messenger RNA in MPM tissue than in normal pleural tissue. The MPM cell lines H2452, H2052, and human mesothelioma cell line MSTO-211H showed a significant increase in expression of AM messenger RNA under hypoxic conditions. Our results also show that AM stimulates cell proliferation in vitro through the Raf1 proto-oncogene, serine/threonine kinase (CRAF)/ Mitogen-activated protein kinase kinase 1 (MEK)/Extracellular regulated MAPKinase (ERK) pathway. Furthermore, the proliferation, migration, and invasion of MPM cells were decreased after treatment with anti-AM (αAM) and anti–AM receptor antibodies, thus indicating that MPM cells are regulated by AM. The action of AM was specific and mediated by CLR/RAMP2 and CLR/RAMP3 receptors. In vivo, αAM and AM22–52 antagonist therapies blocked angiogenesis and induced apoptosis in MSTO-211H xenografts, thereby resulting in tumor regression. Histologic examination of tumors treated with AM22–52 and αAM antibody showed evidence of disruption of tumor vasculature with depletion of vascular endothelial cells and a significant decrease in lymphatic endothelial cells.

      Conclusions

      Our findings highlight the importance of the AM pathway in growth of MPM and in neovascularization by supplying and amplifying signals that are essential for pathologic neoangiogenesis and lymphangiogenesis.

      Keywords

      Introduction

      Human malignant mesothelioma is a tumor of mesothelial origin that is seen most frequently within the lining of the coelomic cavities. Malignant pleural mesothelioma (MPM) is a unique form of mesothelioma, and its development is strongly related to exposure to asbestos fibers.
      • Carbone M.
      • Kratzke R.A.
      • Testa J.R.
      The pathogenesis of mesothelioma.
      However, the mechanism leading to cell transformation in MPM remains largely unknown. Although asbestos use has decreased in many countries in recent years, the incidence of MPM continues to increase worldwide. The increase in cases of MPM can be explained at least partially by the long latency period (30 to 40 years) between exposure to asbestos and tumor development. MPM is an aggressive tumor and notoriously refractory to treatment,
      • Ramalingam S.S.
      • Belani C.P.
      Recent advances in the treatment of malignant pleural mesothelioma.
      and neither surgery nor radiotherapy alone has resulted in increased survival rates in patients with MPM.
      • Ramalingam S.S.
      • Belani C.P.
      Recent advances in the treatment of malignant pleural mesothelioma.
      The prognosis of patients with MPM is extremely poor despite the currently available therapies (such as antifolate pemetrexed combined with cisplatin
      • Vogelzang N.J.
      • Rusthoven J.J.
      • Symanowski J.
      • et al.
      Phase III study of pemetrexed in combination with cisplatin versus cisplatin alone in patients with malignant pleural mesothelioma.
      ), with median survival times varying between 8 and 14 months after diagnosis. Consequently, new therapeutic approaches to treatment of MPM are needed. Solid tumors require angiogenesis, or the development of new blood vessels, for the tumor to grow beyond a minimum size.
      • Folkman J.
      What is the evidence that tumors are angiogenesis dependent?.
      Moreover, many studies have indicated that mesothelioma is a highly angiogenic tumor.
      • Schmitter D.
      • Lauber B.
      • Fagg B.
      • Stahel R.A.
      Hematopoietic growth factors secreted by seven human pleural mesothelioma cell lines: interleukin-6 production as a common feature.
      • Monti G.
      • Jaurand M.C.
      • Monnet I.
      • et al.
      Intrapleural production of interleukin 6 during mesothelioma and its modulation by gamma-interferon treatment.
      • Antony V.B.
      • Hott J.W.
      • Godbey S.W.
      • Holm K.
      Angiogenesis in mesotheliomas. Role of mesothelial cell derived IL-8.
      • Marzo A.L.
      • Fitzpatrick D.R.
      • Robinson B.W.
      • Scott B.
      Antisense oligonucleotides specific for transforming growth factor beta2 inhibit the growth of malignant mesothelioma both in vitro and in vivo.
      • Mutsaers S.E.
      • McAnulty R.J.
      • Laurent G.J.
      • Versnel M.A.
      • Whitaker D.
      • Papadimitriou J.M.
      Cytokine regulation of mesothelial cell proliferation in vitro and in vivo.
      • Kumar-Singh S.
      • Van Marck E.
      Differentiation and differentiating patterns in mesothelioma.
      Thus, the potential efficacy of antiangiogenic therapy for this malignancy, which either shows no response to conventional chemotherapy or rapidly acquires resistance to current treatments, could be particularly high.
      Paracrine and autocrine mechanisms occur in many tumors that simultaneously produce growth factors and express the cognate receptors of these growth factors. Mechanisms such as these are appealing targets in the treatment of many neoplastic diseases.
      • Greig R.
      • Dunnington D.
      • Murthy U.
      • Anzano M.
      Growth factors as novel therapeutic targets in neoplastic disease.
      The adrenomedullin (AM) gene, ADM, is expressed in a wide variety of cell types, including endothelial cells and bone marrow stromal cells, as well as in a variety of tumor cell lines.
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      • De Angeli S.
      • Di Liddo R.
      • Buoro S.
      • et al.
      New immortalized human stromal cell lines enhancing in vitro expansion of cord blood hematopoietic stem cells.
      • Hay D.
      • Walker C.
      • Poyner D.
      Adrenomedullin and calcitonin gene–related peptide receptors in endocrine-related cancers: opportunities and challenges.
      AM is a multifunctional peptide with many properties ranging from the ability to induce vasodilation to the ability to regulate cellular growth and angiogenesis.
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      • Hinson J.P.
      • Kapas S.
      • Smith D.M.
      Adrenomedullin, a multifunctional regulatory peptide.
      • Zhao Y.
      • Hague S.
      • Manek S.
      • Zhang L.
      • Bicknell R.
      • Rees M.C.
      PCR display identifies tamoxifen induction of the novel angiogenic factor adrenomedullin by a nonestrogenic mechanism in the human endometrium.
      AM binds to and mediates its activity through the G protein–coupled receptor calcitonin receptor–like receptor (CLR), with specificity for AM signaling being conferred by receptor activity–modifying protein 2 (RAMP2) and receptor activity–modifying protein 3 (RAMP3).
      • McLatchie L.M.
      • Fraser N.J.
      • Main M.J.
      • et al.
      RAMPs regulate the transport and ligand specificity of the calcitonin-receptor-like receptor.
      The ability of CLR/RAMP2 and CLR/RAMP3 to respond with high affinity to AM implies the existence of two molecularly distinct AM receptors (AMRs), which are referred to as AMR1 and AMR2 receptors.
      • Poyner D.R.
      • Sexton P.M.
      • Marshall I.
      • et al.
      International Union of Pharmacology. XXXII. The mammalian calcitonin gene–related peptides, adrenomedullin, amylin, and calcitonin receptors.
      In addition to their physiological roles, AM and its receptors have been implicated in the pathogenesis of a variety of malignancies. AM is overexpressed in a variety of tumor types,
      • Zudaire E.
      • Martinez A.
      • Cuttitta F.
      Adrenomedullin and cancer.
      and increased expression of AM is associated with disease progression and shortened survival in some patients with cancer.
      • Hay D.
      • Walker C.
      • Poyner D.
      Adrenomedullin and calcitonin gene–related peptide receptors in endocrine-related cancers: opportunities and challenges.
      • Deville J.L.
      • Bartoli C.
      • Berenguer C.
      • et al.
      Expression and role of adrenomedullin in renal tumors and value of its mRNA levels as prognostic factor in clear-cell renal carcinoma.
      • Kocemba K.
      • van Andel H.
      • de Haan-kramer A.
      • et al.
      The hypoxia target adrenomedullin is aberrantly expressed in multiple myeloma and promotes angiogenesis.
      Moreover, AM stimulates proliferation and inhibits apoptosis in a variety of cell types,
      • Hinson J.P.
      • Kapas S.
      • Smith D.M.
      Adrenomedullin, a multifunctional regulatory peptide.
      • Shichiri M.
      • Hirata Y.
      Regulation of cell growth and apoptosis by adrenomedullin.
      • Belloni A.
      • Treijter M.
      • Malendowicz L.
      • Nussdorf G.
      Adrenomedullin stimulates proliferation and inhibits apoptosis of immature rat thymocytes cultured in vitro.
      • Uzan B.
      • Ea H.
      • Launay J.
      • et al.
      A critical role for adrenomedullin-calcitonin receptor-like receptor in regulating rheumatoid fibroblast-like synoviocyte apoptosis.
      including malignant cells from different tumors.
      • Hay D.
      • Walker C.
      • Poyner D.
      Adrenomedullin and calcitonin gene–related peptide receptors in endocrine-related cancers: opportunities and challenges.
      • Deville J.L.
      • Bartoli C.
      • Berenguer C.
      • et al.
      Expression and role of adrenomedullin in renal tumors and value of its mRNA levels as prognostic factor in clear-cell renal carcinoma.
      • Martinez A.
      • Vos M.
      • Guedez L.
      • et al.
      The effects of adrenomedullin overexpression in breast tumor cells.
      • Ouafik L.
      • Sauze S.
      • Boudouresque F.
      • et al.
      Neutralization of adrenomedullin inhibits the growth of human glioblastoma cell lines in vitro and suppresses tumor xenograft growth in vivo.
      • Kaafarani I.
      • Fernandez-Sauze S.
      • Berenguer C.
      • et al.
      Targeting adrenomedullin receptors with systemic delivery of neutralizing antibodies inhibits tumor angiogenesis and suppresses growth of human tumor xenografts in mice.
      AM also promotes the migration and invasiveness of certain carcinoma cell lines.
      • Hay D.
      • Walker C.
      • Poyner D.
      Adrenomedullin and calcitonin gene–related peptide receptors in endocrine-related cancers: opportunities and challenges.
      • Deville J.L.
      • Bartoli C.
      • Berenguer C.
      • et al.
      Expression and role of adrenomedullin in renal tumors and value of its mRNA levels as prognostic factor in clear-cell renal carcinoma.
      • Martinez A.
      • Vos M.
      • Guedez L.
      • et al.
      The effects of adrenomedullin overexpression in breast tumor cells.
      In addition to its effects on malignant cells, AM may promote tumor growth through its ability to promote angiogenesis and lymphangiogenesis.
      • Martinez A.
      • Vos M.
      • Guedez L.
      • et al.
      The effects of adrenomedullin overexpression in breast tumor cells.
      • Ouafik L.
      • Sauze S.
      • Boudouresque F.
      • et al.
      Neutralization of adrenomedullin inhibits the growth of human glioblastoma cell lines in vitro and suppresses tumor xenograft growth in vivo.
      • Kaafarani I.
      • Fernandez-Sauze S.
      • Berenguer C.
      • et al.
      Targeting adrenomedullin receptors with systemic delivery of neutralizing antibodies inhibits tumor angiogenesis and suppresses growth of human tumor xenografts in mice.
      • Berenguer-Daize C.
      • Boudouresque F.
      • Bastide C.
      • et al.
      Adrenomedullin blockade suppresses growth of human hormone-independent prostate tumor xenograft in mice.
      • Oehler M.K.
      • Hague S.
      • Rees M.C.
      • Bicknell R.
      Adrenomedullin promotes formation of xenografted endometrial tumors by stimulation of autocrine growth and angiogenesis.
      • Karpinich N.O.
      • Kechele D.O.
      • Espenschied S.T.
      • Willcockson H.H.
      • Fedoriw Y.
      • Caron K.M.
      Adrenomedullin gene dosage correlates with tumor and lymph node lymphangiogenesis.
      Consistent with this hypothesis, receptor inhibition by the antagonistic peptide AM22–52,
      • Ishikawa T.
      • Chen J.
      • Wang J.
      • et al.
      Adrenomedullin antagonist suppresses in vivo growth of human pancreatic cancer cells in SCID mice by suppressing angiogenesis.
      AM receptor interference,
      • Ramachandran V.
      • Arumugam T.
      • Hwang R.F.
      • Greenson J.K.
      • Simeone D.M.
      • Logsdon C.D.
      Adrenomedullin is expressed in pancreatic cancer and stimulates cell proliferation and invasion in an autocrine manner via the adrenomedullin receptor, ADMR.
      or neutralizing antibodies against AM (αAM) or its receptors reduces the proliferation and invasiveness of tumor-derived cell lines in vitro and reduces the vascularization and growth of experimental tumors in vivo.
      • Ouafik L.
      • Sauze S.
      • Boudouresque F.
      • et al.
      Neutralization of adrenomedullin inhibits the growth of human glioblastoma cell lines in vitro and suppresses tumor xenograft growth in vivo.
      • Kaafarani I.
      • Fernandez-Sauze S.
      • Berenguer C.
      • et al.
      Targeting adrenomedullin receptors with systemic delivery of neutralizing antibodies inhibits tumor angiogenesis and suppresses growth of human tumor xenografts in mice.
      Targeting the AM/AM receptor pathway could be a rational therapeutic approach for treatment of MPM. Because the expression and function of the AM system in MPM has not been well characterized, we evaluated the expression and role of the AM/AM receptor pathway in MPM. Accordingly, the aim of our study was to investigate expression of the AM system (AM and its receptors) and its potential role in the growth of MPM cell lines in vitro and MPM xenografts in vivo.

      Materials and methods

      Human MPM tissues

      We compared specimens of human pleural tissue from patients with MPM (n = 24) who were treated at Assistance Publique Hopitaux de Marseille (AP-HM) (Marseille, France) with specimens of normal pleural tissue (n = 5). Fresh frozen and paraffin-embedded tumor specimens were collected from consenting patients, assigned a deidentifying number, and provided by the AP-HM Tumor Tissue Bank (AC-2013-1786) in accordance with a protocol approved by the relevant institutional committees (AP-HM/Aix-Marseille University). Sections of paraffin-embedded samples (4 μm thick) were analyzed for the presence of AM, CLR, RAMP2, and RAMP3 proteins as previously described,
      • Deville J.L.
      • Bartoli C.
      • Berenguer C.
      • et al.
      Expression and role of adrenomedullin in renal tumors and value of its mRNA levels as prognostic factor in clear-cell renal carcinoma.
      and the proteins' expression was quantified by visual inspection by an experienced pathologist. anti-AM and anti-CLR antibodies were used at dilutions of 1/2000, whereas anti-RAMP2 antibody was used at a dilution of 1/750, and anti-RAMP3 antibody was used at a dilution of 1/1500. As a control for immunostaining, antibodies that had been preabsorbed by human synthetic AM peptide (50 μM; Bachem, Switzerland), CLR, RAMP2 and RAMP3 peptides (50 μM, CROPS laboratory) were used instead of primary antibodies.

      Cell lines and hypoxic treatment

      The cell lines H2452 and H2052 (epithelial mesothelioma), MSTO-211H (biphasic mesothelioma), and MeT-5A (epithelial virus transformed, obtained from mesothelium of noncancerous individuals) were procured from the American Type Culture Collection (Rockville, MD). H2452, H2052, and MSTO-211H cells were cultured in RPMI-1640 (Roswell Park Memorial Institute-1640) medium supplemented with 10% heat-inactivated fetal bovine serum. Met-5A cells were grown in defined medium (M199) as recommended by the American Type Culture Collection. Cells were cultured at 37°C in 20% O2 and 5% CO2 for normoxic conditions. Induction of hypoxia was achieved by using the hypoxia mimetic 260 μM desferrioxamine mesylate (DFX) (Sigma, Paris, France). MPM and MeT-5A cells were grown to 70% confluence. The medium was changed, and the cells were incubated with new medium containing 260 μM DFX for 24 and 48 hours.

      RNA preparation and real-time quantitative reverse transcriptase-polymerase chain reaction

      Before preparation of the RNA, a cryostat section was taken from each specimen to confirm the presence of tumor tissue. Total RNA was prepared from fresh frozen biopsy samples of MPM tumors (n = 24), normal pleural tissue (n = 5), and H2452, H2052, MSTO-211H, and MeT-5A cells; it was then reverse transcribed to cDNA as described previously.
      • Berenguer C.
      • Boudouresque F.
      • Dussert C.
      • et al.
      Adrenomedullin, an autocrine/paracrine factor induced by androgen withdrawal, stimulates “neuroendocrine phenotype” in LNCaP prostate tumor cells.
      Human AM, CLR, RAMP2, RAMP3, and glyceraldehyde-3-phosphate degydrogenase (GAPDH) messenger RNAs (mRNAs) were amplified, detected, and quantified in real time by using an LC 480 polymerase chain reaction (PCR) system (Roche Diagnostics, Meylan, France) as described previously.
      • Deville J.L.
      • Bartoli C.
      • Berenguer C.
      • et al.
      Expression and role of adrenomedullin in renal tumors and value of its mRNA levels as prognostic factor in clear-cell renal carcinoma.
      • Berenguer C.
      • Boudouresque F.
      • Dussert C.
      • et al.
      Adrenomedullin, an autocrine/paracrine factor induced by androgen withdrawal, stimulates “neuroendocrine phenotype” in LNCaP prostate tumor cells.

      In vitro capillary tube formation on Matrigel

      The morphogenesis assay on Matrigel was performed as described previously.
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      Human umbilical vein endothelial cells (HUVECs) maintained in EBM2 medium with supplement were washed twice with phosphate-buffered saline (PBS), trypsinized, and plated (4 × 104 cells) in wells coated with growth factor–depleted Matrigel (7mg/mL; Becton Dickinson, Paris, France) in normoxia or hypoxia medium of H2452 or MSTO-211H with 0.5% fetal bovine serum alone or in presence of anti-AM (αAM) and anti-AMR (αAMR) antibodies. After 2 or 4 hours, the plates were photographed and the extent of tube formation was assessed qualitatively.

      Western blot analysis

      Cell extracts were prepared and immunoblotted for phospho-Raf1 proto-oncogene, serine/threonine kinase (pCRAF), phospho-mitogen-activated protein kinase kinase1 (pMEK1), phospho extracellular regulated MAPKinase (pERK1/2), and extracellular regulated MAPKinase (ERK1/2) using the mitogen-activated protein kinase (MAPK)-phospho-ERK1/2 pathway sampler kit (Cell Signaling Technology Inc.) as described previously.
      • Berenguer-Daize C.
      • Boudouresque F.
      • Bastide C.
      • et al.
      Adrenomedullin blockade suppresses growth of human hormone-independent prostate tumor xenograft in mice.
      Phospho-p38 was detected with monoclonal antibody (Cell Signaling Technology Inc.). Antibody signals were revealed by using an enhanced chemiluminescence kit (ECL Kit, Invitrogen Life Technologies Inc.).

      Cell proliferation assay

      The effects of AM (10–7 M), AM22–52 (10–6 M), rabbit antihuman αAM (70 μg/mL), and antihuman AMR (αCLR, αRAMP2, and αRAMP3; 70 μg/mL)-neutralizing antibodies (purified immunoglobulin G [IgG]), which were previously developed in house,
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      • Ouafik L.
      • Sauze S.
      • Boudouresque F.
      • et al.
      Neutralization of adrenomedullin inhibits the growth of human glioblastoma cell lines in vitro and suppresses tumor xenograft growth in vivo.
      on cell proliferation were examined by cell counting (Z1 Series Coulter Counter, Beckman Coulter Inc., Fullerton, CA).

      Cell migration and invasion assays

      A modified Boyden chamber assay was used to analyze migration and chemoinvasion of H2452 and MSTO-211H cells as described previously.
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      • Deville J.L.
      • Bartoli C.
      • Berenguer C.
      • et al.
      Expression and role of adrenomedullin in renal tumors and value of its mRNA levels as prognostic factor in clear-cell renal carcinoma.
      • Albini A.
      • Iwamoto Y.
      • Kleinman H.K.
      • et al.
      A rapid in vitro assay for quantitating the invasive potential of tumor cells.

      Animal experiments

      Athymic naval medical research institute (NMRI; nu/nu) nude mice (Harlan Laboratories SARL, France) were purchased at 5 weeks of age. Suspensions of MSTO-211H cells (2 × 106 in 100 μL PBS) were injected subcutaneously into the right flank of male mice (n = 30). Tumor sizes were determined by dial caliper measurements, and tumor volumes were calculated with the equation for an ellipsoid form: width × length × height × 0.5236. Mice with subcutaneous (SC) tumors larger than 1.5 cm in diameter were humanely killed in accordance with Aix-Marseille University Animal Rights Committee guidelines. When the tumors reached 250 ± 50 mm3, the animals were randomly divided into two groups. One group (n = 10) received an intraperitoneal (IP) injection of αAM (12 mg/kg purified IgG in 200 μL PBS) every 3 days, and the second group (n = 10) received 50 μg of the AM antagonist AM22–52 daily as described previously.
      • Ishikawa T.
      • Chen J.
      • Wang J.
      • et al.
      Adrenomedullin antagonist suppresses in vivo growth of human pancreatic cancer cells in SCID mice by suppressing angiogenesis.
      αAM antibody was characterized as described previously,
      • Berenguer-Daize C.
      • Boudouresque F.
      • Bastide C.
      • et al.
      Adrenomedullin blockade suppresses growth of human hormone-independent prostate tumor xenograft in mice.
      and all IgG preparations were tested for endotoxin using the Pyrogent Plus Limilus Amebocyte Lysate kit (Lonza). All antibody preparations used in our animal studies contained less than 1.25 U/mL endotoxin. A control group (n = 10) received an irrelevant antibody (IgG of the same isotype). Tumor sizes were measured every 3 days, and the mice were humanely killed 7 weeks after tumor implantation. Tumors were stored in liquid nitrogen or embedded in paraffin for pathologic analysis and immunohistochemistry.

      Immunohistochemical staining

      Sections (6 μm thick) were cut from formalin-fixed paraffin-embedded MSTO-211H xenografts. Immunohistochemistry was performed using the Vectastatin Elite ABC Universal kit (Vector Laboratories, Burlingame, CA) as described previously.
      • Deville J.L.
      • Bartoli C.
      • Berenguer C.
      • et al.
      Expression and role of adrenomedullin in renal tumors and value of its mRNA levels as prognostic factor in clear-cell renal carcinoma.
      • Berenguer-Daize C.
      • Boudouresque F.
      • Bastide C.
      • et al.
      Adrenomedullin blockade suppresses growth of human hormone-independent prostate tumor xenograft in mice.
      Antibodies recognizing CD31 (1:20, Dianova), Ki67 nuclear antigen (1:100; Dako), cleaved caspase-3 (1:100; BD Pharmingen), and lymphatic vessel hyaluronic acid receptor-1 (LYVE-1) (1:100) were used for our analysis. For each marker, whole-surface staining was quantified using CALOPIX Software (TRIBVIN Medical, Châtillon, France).

      Statistical analysis

      Data are expressed as the means ± standard error of the mean (SEM) from at least three independent experiments. One-way analysis of variance or Fisher’s partial least squares difference test (Statview 512; Brain Power Inc., Calabasas, CA) was used for statistical analysis. Differences were considered significant at values of p < 0.05.

      Results

      Expression of AM, CLR, RAMP2, and RAMP3 mRNAs in human MPM

      Total RNA from normal pleural tissue (n = 5) and MPM (n = 24) was prepared to assess the steady-state levels of AM, CLR, RAMP2, and RAMP3 mRNA transcripts. The individual patterns of expression of AM mRNA are presented in Fig. 1A. Quantification of the AM mRNA transcripts revealed twofold to 10-fold higher levels of AM mRNA in MPM tissue (≈20 to 110 fg/pg GAPDH mRNA) than in normal pleural tissue (≈10 fg/pg GAPDH mRNA). Among the MPM samples, the individual pattern of expression for AM, CLR, RAMP2, and RAMP3 mRNAs was highly variable (see Fig. 1A). MPM tissues also showed levels of expression of CLR mRNA ranging between 0.01 and 0.06 fg/pg GAPDH mRNA, RAMP2 mRNA levels between 0.1 and 1.9 fg/pg GAPDH mRNA, and RAMP3 mRNA levels between 0.01 to 0.5 fg/pg GAPDH mRNA, compared with mRNAs levels between 0.0078 and 0.012 fg/pg GAPDH mRNA, between 0.075 and 0.144 fg/pg GAPDH mRNA, and between 0.005 and 0.01fg/pg GAPDH mRNA in normal pleura tissue for CLR, RAMP2, and RAMP3, respectively (see Fig. 1A). Omission of reverse transcriptase eliminated all signals, thus suggesting that our results were not attributable to contaminating genomic DNA (data not shown).
      Figure thumbnail gr1
      Figure 1Expression of adrenomedullin (AM) and its receptors in human malignant pleural mesothelioma (MPM). (A) Expression of AM and its receptors (calcitonin receptor–like receptor [CLR], receptor activity–modifying protein 2 [RAMP2], and receptor activity–modifying protein 3 [RAMP3]) mRNAs in normal pleura and MPM tissues. Total RNA (1 μg, DNA-free) prepared from normal pleural tissue (n = 5) and MPM tissue (n = 24) was transcribed into cDNA and subjected to real-time quantitative reverse transcriptase-polymerase chain reaction for the estimation of the relative ratios of AM, CLR, RAMP2, and RAMP3 mRNAs to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA as described in Materials and Methods. Each bar depicts the mean ± standard error of the mean of two independent experiments from two independent preparations of total RNA from MPM tissues. (B) Immunohistochemistry for AM, CLR, RAMP2, and RAMP3 in epithelial MPM. Strong cytoplasmic and nuclear staining for AM, CLR, RAMP2, and RAMP3 is observed in MPM cells (red arrows). Stroma cells with weaker cytoplasmic staining are also observed (blue arrows).

      Immunohistochemistry of AM, CLR, RAMP2, and RAMP3 proteins in human MPM

      Sections of epithelial MPM were labeled with antibodies revealing AM, CLR, RAMP2, and RAMP3 proteins (Fig. 1B). MPM cells displayed overt and strong labeling for AM, CLR, RAMP2, and RAMP3 (see Fig. 1B). Illustrating the complexity of their localization, the immunostaining revealed strong cytoplasmic and nuclear staining for AM, CLR, and RAMP2 in epithelial mesothelioma (see Fig. 1B, red arrows). The same localization for AM was recently reported in serial sections of lung cancer biopsy samples and prostate cancer specimens.
      • Berenguer-Daize C.
      • Boudouresque F.
      • Bastide C.
      • et al.
      Adrenomedullin blockade suppresses growth of human hormone-independent prostate tumor xenograft in mice.
      Numerous clusters of labeled stromal cells with weaker labeling for AM, CLR, and RAMP2 could be observed dispersed among the stromal collagen septa, thus suggesting that AM system is potentially involved in tumor–stroma cross talk (see Fig. 1B; blue arrows). Positive AM, CLR, RAMP2, and RAMP3 staining was completely abolished when antibodies were preabsorbed with 50 μmol/L synthetic AM, CLR, RAMP2, and RAMP3 peptides (data not shown).

      Expression of AM and AM receptors in MPM cells

      The expression of AM and AM receptors in MPM suggests that the AM system plays a role in promoting tumor growth in situ. Accordingly, we used H2452, H2052, MSTO-211H, and MeT-5A cells to gain insight into the role of the AM system in MPM cells. The presence and cellular localization of AM and AM receptors in cell lines under normoxic conditions were analyzed using immunofluorescence. Representative images are shown in Fig. 2A; in the images H2452 and MSTO-211H cells have been immunostained for AM, CLR, RAMP2, and RAMP3 (results concerning H2052 and MeT-5A cells are shown in Supplementary Fig. 1A). Under normoxic conditions, AM, CLR, and RAMP2 staining generally localized to the cytoplasm, except in H2452 cells, in which AM staining localized to the nucleus (Fig. 2A).
      Figure thumbnail gr2
      Figure 2Expression and regulation of adrenomedullin (AM) signaling in malignant pleural mesothelioma (MPM) cells. (A) Expression of AM and its receptors in MPM cells. Immunofluorescence of H2452 and MSTO-211H cells stained with antibodies against AM, calcitonin receptor–like receptor, receptor activity–modifying protein 2, and receptor activity–modifying protein 3 revealed that except in H2452 cells (which demonstrated nuclear localization for AM), these proteins localized to the cytoplasm. (B) Hypoxia induces the expression of AM in MPM cells. Total RNA (1 μg, DNA-free) prepared from H2452 and MSTO-211H cells under normoxic and hypoxic conditions were reverse transcribed into cDNA and subjected to quantitative reverse transcriptase polymerase chain reaction for the estimation of the relative AM mRNA. Each bar represents the mean ± standard error of the mean of three independent experiments from three independent preparations of total RNA. Significant differences in the expression of AM in desferrioxamine mesylate (DFX)-treated cells with respect to untreated controls were determined by a one-way analysis of variance test (***p < 0.001). (C) Cells were grown on glass slides and then incubated under normoxic conditions or treated with 260 μM DFX for 24 hours, after which the cells were immunostained for AM (green fluorescence). Identical microscope settings and exposure times were used for the images obtained for normoxic and DFX-treated cells to compare the expression levels under both conditions. AM was markedly overexpressed after 24 hours of DFX treatment. (D) Morphogenic activity of hypoxia medium. Human umbilical vein endothelial cells (4 × 104 cells/well) were seeded into Matrigel-precoated wells and cultured in low-serum conditions (0.5% fetal calf serum) in normoxia MSTO-211H medium for 24 hours (a), in hypoxia MSTO-211H medium alone for 24 hours (b), or in presence of 20 μg/mL anti-AM (c) or 20 μg/mL anti–AM receptor (d) antibodies. Photographs were taken 4 hours later.

      Regulation of AM expression by hypoxia

      Quantitative reverse transcriptase PCR (RT-PCR) analysis demonstrated that H2452 and MSTO-211H cells express AM mRNA (Fig. 2B). Under hypoxic conditions, the levels of AM mRNA increased threefold to fourfold in H2452 cells (Fig. 2B), twofold to 2.5-fold in MSTO-211H cells (Fig. 2B), and threefold to fivefold in H2052 cells (Supplementary Fig. 1B) as early as after 24 hours. The MeT-5A cells showed a sevenfold increase of AM mRNA levels only after 48 hours of incubation with hypoxia medium (see Supplementary Fig. 1B). To determine whether the induction of AM mRNA was accompanied by an increase in production of AM protein, the presence and cellular localization of AM in tumor cell lines under normoxic or hypoxic conditions was determined by immunofluorescence. A representative image is shown in Figure 2C, in which MPM cells H2452 and MSTO-211H have been immunostained for AM (green fluorescence) using highly diluted anti-AM antibody (1/5000). At this dilution, the immunoreactive AM was barely detectable in the cytoplasm and nucleus of the cells under normoxic conditions (see Fig. 2C). After 24 hours of exposure to 260 μM DFX, however, the cells showed a marked increase in AM staining in the cytoplasm and nucleus (see Fig. 2C), which is consistent with the up-regulation of AM synthesis under hypoxic conditions. The production and secretion of AM by hypoxic cells may have an angiogenic effect by promoting a stable and functional vascular network necessary for tumor growth in vivo. To strengthen our findings, we explored the ability of AM secreted by MSTO-211H and H2452 cells to induce angiogenesis and tested this hypothesis in an in vitro Matrigel assay. Because AM receptors are expressed in cultured primary HUVECs,
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      the latter might respond to AM secreted by MSTO-211H and H2452 cells under hypoxia to accelerate the formation of vascular structures. The addition of medium secreted by MSTO-211H cells cultured under hypoxia for 24 hours resulted in a marked increase in morphogenesis in vitro compared with that observed after 24 hours in a normoxia medium (Fig. 2D, a and b); the observed morphogenesis was characterized by cell elongation and branching to form a network of capillary-like structures. This effect was inhibited by function-blocking antibodies to AM and AM receptors (see Fig. 2D, c and d). The same data were obtained with HUVECs incubated with the hypoxia medium of H2452 cells for 24 hours (Supplementary Fig. S2). These data strongly suggest that AM is one of the MPM cell–derived factors secreted under hypoxia that are responsible for inducing angiogenesis in vivo.

      Effects of AM and AM blockade on MPM cell proliferation

      The expression of AM and its receptors in MPM tissues and in cell lines suggests that AM may be involved in MPM cell growth through an autocrine/paracrine growth loop. Of the three lines of MPM cells (MSTO-211H, H2452, and H2052), the MSTO-211H cells showed an approximately 20% increase in proliferation in the presence of the maximum AM concentrations (10–7 M) when compared with untreated cells after 6 days of treatment (Fig. 3A). Consistent with an autocrine function for AM in these cells, αAM- or αAMR-neutralizing antibodies added to MSTO-211H, H2452, and H2052 cell culture medium significantly reduced cell proliferation by as much as 40% for MSTO-211H cells (p < 0.01), 30% for H2452 cells (p < 0.01), and 30% to 40% for H2052 cells (p < 0.01; p < 0.001) when compared with cells treated with a nonspecific, isotype control antibody (see Fig. 3A). MeT-5A cells demonstrated a lower but significant decrease in proliferation when incubated with αAM and αAMR antibodies (p < 0.05) (see Fig. 3A). AM22–52 reduced proliferation by up to 35% in MSTO-211H cells, whereas treatment with AM22–52 had no effect on H2452, H2052, and MeT-5A cell proliferation (see Fig. 3A). Taken together, these observations indicate that AM is involved in the cell growth, notably in the three MPM cell lines, through AMR1 and AMR2.
      Figure thumbnail gr3
      Figure 3Effect of adrenomedullin (AM) and AM signaling blockade on growth, migration, and invasion of mesothelioma cells in vitro. (A) For the proliferation assays, cells were seeded at a density of 2 × 103 cells per well in 24 multiwell plates in the presence of medium containing 2% fetal bovine serum. AM (10–7 M), anti-AM antibody (αAM) (70 μg/mL), anti–AM receptor antibody (αAMR) (70 μg/mL), AM22–52 peptide (10–6 M), or control immunoglobulin G (70 μg/mL) was added to the cells for 6 days of treatment. For each treatment, six wells were prepared for 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) analysis. Each bar represents the mean ± standard error of the mean of three independent experiments. Significant differences between the growth of cells treated with αAM, αAMR, and AM22–52 and that of untreated controls were determined by a one-way analysis of variance test (*p < 0.05; **p < 0.01; ***p < 0.001). (B and C), AM regulates MPM cell migration and invasion. The bottom wells of all chambers were filled with RPMI-1640 (Roswell Park Memorial Institute-1640) medium containing 2% fetal bovine serum in the presence of control buffer (control), AM (10–7 M), or endothelial growth factor (10–7 M). MSTO-211H cells (B, 5 × 104 cells) or H2452 cells (C, 2 × 104 cells) pretreated for 30 minutes with αAM (70 μg/mL), αCLR (23 μg/mL), anti–receptor activity–modifying protein-2 (23 μg/mL), anti–receptor activity–modifying protein-3 (23 μg/mL, αAMR), or control immunoglobulin G (70 μg/mL) were placed in the upper chamber and incubated for 4 hours with MSTO-211H cells (B) or 16 hours with H2452 cells (C) at 37°C. The cells that migrated were stained with 4’,6’-diamidino-2-phenylindole and counted at 50× magnification using a microscope. Data are expressed as the number of migrated cells in 10 high-power fields, and the values represent the mean ± standard error of the mean of three independent experiments, each performed in triplicate. Significant differences in the number of treated cells with respect to untreated controls were detected by a one-way analysis of variance test and were observed in the two cell lines. The asterisk (*) is used for comparison to control cells (*p < 0.05; **p < 0.01; ***p < 0.001) and the plus symbol (+) is used in comparison to AM-treated cells (++p < 0.01; +++p < 0.001).

      AM induces MPM cell migration and invasion in vitro

      To determine whether AM could affect MPM cell motility, MSTO-211H and H2452 cells were incubated for 4 and 16 hours, respectively, in a Boyden chamber assay in the presence of AM. The number of cells that migrated to the lower surface of the transwell apparatus was assessed under different conditions (Figs. 3B and C). In the presence of AM (10–7 M), no significant stimulation was observed in MSTO-211H (see Fig. 3B), whereas a highly significant increase in migration was observed for H2452 cells (p < 0.01) (see Fig. 3C). Addition of AM (10–7 M) showed no significant stimulation of invasion of H2452 and MSTO-211H cells (see Figs. 3B and C). Preincubation of either MSTO-211H or H2452 cells with a combination of αCLR, αRAMP2, and αRAMP3 or αAM for 30 minutes inhibited the migration of H2452 cells but not MSTO-211H cells (see Figs. 3B and C). However, invasion was completely impaired for both cell types (see Figs. 3B and C), indicating that AM promotes cell invasion through AMR1 and AMR2. Taken together, our data demonstrate the presence of an autocrine loop that involves AM receptors and secreted AM, which supports the migration and invasion of MPM cells (see Figs. 3B and C). The addition of preimmune control IgG at 70 μg/mL did not affect the stimulatory action of AM on the MPM cell lines (see Figs. 3B and C). Treatment with endothelial growth factor was used as positive control.

      AM mediates the phosphorylation of MAPK

      ERK and serine/threonine protein kinase (Akt) regulate cell proliferation, and both of these signaling pathways function downstream of the AM/cyclic adenosine monophosphate pathway.
      • Ilan N.
      • Mahooti S.
      • Madri J.A.
      Distinct signal transduction pathways are utilized during the tube formation and survival phases of in vitro angiogenesis.
      • Yu Y.
      • Sato J.D.
      MAP kinases, phosphatidylinositol 3-kinase, and p70 S6 kinase mediate the mitogenic response of human endothelial cells to vascular endothelial growth factor.
      Therefore, we examined the kinetics through which AM enhanced MAPK signaling. Treatment of MSTO-211H cells and H2452 cells with AM led to prolonged phosphorylation of both CRAF and MEK that was initially observed after 5 minutes of treatment and continued to increase through 120 minutes of treatment (Figs. 4A and B). Of note, the activation of MEK by AM was more apparent in H2452 cells than in MSTO-211H cells, in which a very high constitutive activation of MEK occurred. These increases in signaling correlated with the sustained phosphorylation of ERK1/2 MAPK in MSTO-211H and H2452 cells (see Figs. 4A and B), as well as in H2052 cells (Supplementary Fig. 3). Levels of pCRAF in H2452 cells declined more rapidly after 20 minutes of treatment (see Fig. 4B). Inhibition of MEK, an immediate upstream activator of ERK1/2, with U0126 (10 μM for 30 minutes) prevented AM-mediated activation of ERK1/2 (see Figs. 4A and B and Supplementary Fig. S3). No induction of pERK1/2 by AM could be observed in MeT-5A cells (see Supplementary Fig. S3). Preincubation of H2452 and MSTO-211H cells with αAM, αAMR, αCLR, αRAMP2, αRAMP3, αCLR/αRAMP2, and αCLR/αRAMP3 decreased the stimulatory effect of AM on pERK1/2 (see Figs. 4A and B). Although a greater decrease was observed with αCLR/αRAMP3, it seems that both complexes, αCLR/αRAMP2 (AMR1) and αCLR/αRAMP3 (AMR2), are involved in the signaling activated by AM in MPM cells (see Figs. 4A and B). Treatment with AM had a strong effect on the phosphorylation of p38, which peaked at 5 minutes and declined at 30 minutes in MSTO-211H cells (see Fig. 4A) and at 1 hour in H2452 cells (see Fig. 4B). These data suggest that AM enhances and prolongs activation of the MAPK pathway through AMR1 and AMR2.
      Figure thumbnail gr4
      Figure 4Intracellular signaling pathway induced by adrenomedullin (AM) in malignant pleural mesothelioma (MPM) cells. MSTO-211H cells (A) and H2452 cells (B) were treated with AM (10–7 M) for the indicated amounts of time and then immunoblotted for phospho-CRAF, CRAF, pMEK1/2, pERK1/2, ERK1/2, and pP38. MEK inhibitor (U0126) inhibited AM-induced phosphorylation of ERK (10 μmol/L, 30 minutes). Endothelial growth factor was used as a positive control to stimulate the phosphorylation of CRAF, MEK1/2, and ERK1/2, and P38. Preincubation of H2452 and MSTO-211H cells with anti-AM, anti–AM receptor, anti–calcitonin receptor–like receptor (αCLR), anti–receptor activity–modifying protein 2 (αRAMP2), anti–receptor activity–modifying protein 3 (αRAMP3), αCLR/αRAMP2, or αCLR/αRAMP3 for 30 minutes decreased AM-induced phosphorylation of ERK1/2. β-Actin was used as a loading control.
      CRAF, Raf1 proto-oncogene, serin/threonine kinase; MEK, Mitogen-activated protein kinase kinase 1; ERK, Extracellular regulated MAPKinase.

      AM blockade inhibits the growth of MSTO-211H tumor xenografts in vivo

      To assess the expression and source of immunoreactive AM (ir-AM) in nude mice xenografted with SC MSTO-211H cells, tissue sections (5 μm) were prepared from SC xenografts, brain, and kidney from the same nude mice and immunostained for ir-AM. The results showed a high level of expression of ir-AM in MSTO-211H xenografts but no expression in brain and kidney (Fig. 5A), thus suggesting that the xenografts were the major source of human AM secreted in the animals. Positive AM staining was completely abolished by preabsorption of the antibody with 50 μmol/L synthetic AM peptide (see Fig. 5A).
      Figure thumbnail gr5
      Figure 5Adrenomedullin (AM) signaling blockade inhibited the growth of MSTO-211H xenografts in vivo. (A) Immunohistochemistry for immunoreactive AM in the MSTO-211H xenografts, brain, and kidney of nude mice. Strong staining was observed in the xenograft specimens. In the brain and kidney, immunoreactive AM was undetectable. (B) MSTO-211H cells (2 × 106) were injected subcutaneously into the flanks of athymic nude mice (6 weeks old) (n = 10 in each group). Mice with tumor volumes averaging 200 to 300 mm3 received intraperitoneal injections of anti-AM (αAM) (12 mg/kg) every 3 days or AM22–52 peptide (50 μg/mouse) daily. Control mice were treated with 12 mg/kg of nonspecific isotype control immunoglobulin G (IgG). Tumor size was measured every 3 days, and significant differences between the animals treated with αAM- and AM22–52 and the animals treated with control IgG were determined by a one-way analysis of variance test (*p < 0.05; **p < 0.01). (C) Tumors were weighed immediately after excision. The average tumor weight is indicated as the mean ± standard error of the mean (n = 10). (D) The photograph shows an example of tumors from the animals treated with αAM and control IgG that were collected at the end of treatments. Note the pale character and the apparent diminished vasculature in tumor from the αAM-treated animal when compared with tumor from control IgG–treated animals.
      To evaluate the functional role of AM in MPM tumorigenesis, we investigated the effects of inhibition of AM signaling on tumor xenografts. To assess the potential therapeutic value of αAM antibody and AM22–52, athymic nude mice bearing established MSTO-211H tumor xenografts (>200 mm3) were treated with αAM, AM22–52, or a rabbit control IgG. Treatment was administered by IP injection every day for AM22–52 (50 μg/mouse), as reported previously,
      • Ishikawa T.
      • Chen J.
      • Wang J.
      • et al.
      Adrenomedullin antagonist suppresses in vivo growth of human pancreatic cancer cells in SCID mice by suppressing angiogenesis.
      and every 3 days for αAM antibody (12 mg/kg) and control IgG (12 mg/kg). To monitor tumor growth, tumor volume was measured throughout the treatment period. The growth of the MSTO-211H xenografts was significantly inhibited by αAM and AM22–52 treatments compared with that in the control group (Fig. 5B). After a 24-day treatment period, a group of animals (n = 5) was humanely killed, and tumor size and vascularity were assessed. The mean tumor weights in the animals treated with control IgG, αAM antibody, and AM22–52 were 2.80 ± 0.40 g, 1.35 ± 0.25 g, and 1.55 ± 0.20 g, respectively, after 27 days of treatment (Fig. 5C). Tumors from the animals treated with αAM antibody and AM22–52 were pale and showed diminished vasculature, whereas large tumors with extensive vascularization were observed in the control groups (Fig. 5D).

      AM blockade impairs tumor angiogenesis and lymphangiogenesis and induces apoptosis

      Histopathologic analysis revealed that large necrotic areas were more prevalent in the tumors from animals treated with αAM antibody and AM22–52 than in the tumors from the animals treated with control IgG (Fig. 6A). Immunohistochemical staining performed on the tumor xenografts demonstrated no significant differences in the Ki-67 labeling index between the animals treated with αAM antibody, AM22–52, and control IgG (Figs. 6A and 6B). Significantly higher numbers of cleaved caspase-3–positive cells were observed in tumors from the animals treated with αAM antibody and AM22–52 (Fig. 6C). Consistent with our hypothesis that AM signaling inhibition would result in a decrease in angiogenesis and lymphangiogenesis, immunohistochemical staining for the endothelial cell marker CD31 and the lymphatic endothelial cell marker LYVE-1 demonstrated that tumor vascularization is deeply disrupted in tumors from animals treated with αAM antibody and AM22–52 (see Fig. 6A). Quantification of CD31-positive endothelial cells and LYVE-1–positive lymphatic endothelial cells demonstrated a clear decrease in both cell types in tumors from animals treated with αAM antibody and AM22–52 compared with the levels in control MSTO-211H tumors (p < 0.1; p < 0.001; Figs. 6A, D, and E).
      Figure thumbnail gr6
      Figure 6Blockade of adrenomedullin (AM) signaling induces apoptosis and impairs angiogenesis and lymphangiogenesis in malignant pleural mesothelioma tumor xenografts. (A) Representative images of tumors from the animals treated with control immunoglobulin G, AM22–52, and anti-AM antibody. Tumor sections were stained with hematoxylin and eosin, Ki-67, cleaved caspase-3, CD-31, and lymphatic vessel hyaluronic acid receptor-1 (LYVE-1) antibodies. Necrotic areas surrounding the tumor tissue are clearly shown in the tumors treated with AM22–52 and anti-AM antibody. Cleaved caspase-3– and Ki-67–positive cells are shown; they were analyzed on the basis of 10 magnification fields (400×) per section. Immunohistochemical staining of the endothelial cell surface marker CD31 was used to determined microvessel density. Quantitative assessment of the density of cells that stained positive for Ki-67 (B), cleaved caspase-3 (C), CD-31 (D), or LYVE-1 (E) was conducted for the entire surface of the corresponding slides using CALOPIX Software. MBF_Image J 1.43U software was used for the analysis. The values shown represent the means ± standard error of the mean (*p < 0.05; **p < 0.01; ***p < 0.001; n.s., nonsignificant).

      Discussion

      The complex nature of MPM is reflected in the lack of success of the current therapeutic strategies. Regardless of the therapy, a dramatic difference in survival between treated and untreated patients with MPM has not been observed.
      • Scherpereel A.
      • Astoul P.
      • Baas P.
      • et al.
      Guidelines of the European Respiratory Society and the European Society of Thoracic Surgeons for the management of malignant pleural mesothelioma.
      Therefore, research aimed at defining the growth mechanisms in MPM is imperative for identifying relevant biological targets and designing new clinical trials in an informed manner. In this study, we investigated the expression of AM and its receptors in pleural tumors from patients with MPM and the potential role of AM as an autocrine/paracrine growth factor to promote growth of MPM in vitro and in vivo. Real-time quantitative RT-PCR results revealed a significant increase in the expression of AM mRNA in MPM samples compared with that in normal pleura. Moreover, immunohistochemical analysis revealed that AM and AM receptors localize mainly to MPM cells. The expression of AM and its receptors in MPM specimens strongly suggests that AM could be involved in progression of MPM.
      Our data demonstrate that hypoxic conditions induce a severalfold increase in AM expression in MSTO-211H, H2452, and H2052 cells. The increase in production and secretion of bioactive AM in these cells under hypoxic conditions suggests that a similar scenario may occur in tissue neoplasms. The production and secretion of AM at hypoxic areas present in tumors
      • Brown J.M.
      • Giaccia A.J.
      The unique physiology of solid tumors: opportunities (and problems) for cancer therapy.
      • Metellus P.
      • Voutsinos-Porche B.
      • Nanni-Metellus I.
      • et al.
      Adrenomedullin expression and regulation in human glioblastoma, cultured human glioblastoma cell lines and pilocytic astrocytoma.
      could establish an autocrine/paracrine-mediated proliferation event leading to tumor growth. In addition, the angiogenic and vasodilator capabilities of AM
      • Zhao Y.
      • Hague S.
      • Manek S.
      • Zhang L.
      • Bicknell R.
      • Rees M.C.
      PCR display identifies tamoxifen induction of the novel angiogenic factor adrenomedullin by a nonestrogenic mechanism in the human endometrium.
      • Kitamura K.
      • Kangawa K.
      • Kawamoto M.
      • et al.
      Adrenomedullin: a novel hypotensive peptide isolated from human pheochromocytoma.
      may be associated with its ability to promote a stable and functional vascular network and facilitate the nutritional supplementation of tumor cells. Finally, AM protects cells from apoptosis,
      • Kato H.
      • Shichiri M.
      • Manumo F.
      • et al.
      Adrenomedullin as an autocrine/paracrine apoptosis survival factor for rat endothelial cells.
      and this feature could selectively rescue tumor cells from death, thereby resulting in the predisposition of tumors toward a more malignant phenotype.
      • Brown J.M.
      • Giaccia A.J.
      The unique physiology of solid tumors: opportunities (and problems) for cancer therapy.
      The presence of AM, AMR1, and AMR2 in MPM tissue suggests that AM may act as an autocrine/paracrine growth factor in MPM tumors. Our data show that AM significantly increases proliferation and invasiveness in MSTO-211H and H2452 cells and increases activation of the CRAF/MEK/ERK/MAPK pathway. These data indicate that MPM cells respond to AM in a manner that is consistent with an increase in the aggressiveness of MPM. We demonstrated that αAM and αAMR antibodies could inhibit the basal levels of MPM cell proliferation and invasion in vitro, which is consistent with the fact that AM can act in an autocrine manner in MPM cells. The presence of an autocrine loop suggests that the foci of AM-producing cells in tumor tissue could stimulate cells expressing AM receptors through autocrine/paracrine mechanisms. Our results suggesting that AM can act as an autocrine/paracrine growth factor in MPM tissue are in agreement with previously reported data on other forms of cancer in which AM acts in a similar autocrine/paracrine manner.
      • Greig R.
      • Dunnington D.
      • Murthy U.
      • Anzano M.
      Growth factors as novel therapeutic targets in neoplastic disease.
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      • Hinson J.P.
      • Kapas S.
      • Smith D.M.
      Adrenomedullin, a multifunctional regulatory peptide.
      • Zhao Y.
      • Hague S.
      • Manek S.
      • Zhang L.
      • Bicknell R.
      • Rees M.C.
      PCR display identifies tamoxifen induction of the novel angiogenic factor adrenomedullin by a nonestrogenic mechanism in the human endometrium.
      • McLatchie L.M.
      • Fraser N.J.
      • Main M.J.
      • et al.
      RAMPs regulate the transport and ligand specificity of the calcitonin-receptor-like receptor.
      • Poyner D.R.
      • Sexton P.M.
      • Marshall I.
      • et al.
      International Union of Pharmacology. XXXII. The mammalian calcitonin gene–related peptides, adrenomedullin, amylin, and calcitonin receptors.
      • Zudaire E.
      • Martinez A.
      • Cuttitta F.
      Adrenomedullin and cancer.
      The role of AM in the stimulation of tumor cell proliferation, inhibition of apoptosis, and stabilization of angiogenesis has been observed in several malignancies,
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      • Ouafik L.
      • Sauze S.
      • Boudouresque F.
      • et al.
      Neutralization of adrenomedullin inhibits the growth of human glioblastoma cell lines in vitro and suppresses tumor xenograft growth in vivo.
      • Kaafarani I.
      • Fernandez-Sauze S.
      • Berenguer C.
      • et al.
      Targeting adrenomedullin receptors with systemic delivery of neutralizing antibodies inhibits tumor angiogenesis and suppresses growth of human tumor xenografts in mice.
      • Oehler M.K.
      • Hague S.
      • Rees M.C.
      • Bicknell R.
      Adrenomedullin promotes formation of xenografted endometrial tumors by stimulation of autocrine growth and angiogenesis.
      and these activities may be relevant in the development of MPM. Our results demonstrate that the blockade of AM signaling by αAM antibody or AM antagonist peptide (AM22–52) significantly decreases the growth of established MSTO-211H tumor xenografts. After 27 days of treatment, the tumors in mice treated with control IgG grew rapidly to a size requiring humane killing of the animal, whereas the volume of the tumors of the animals treated with αAM antibody or AM22–52 remained significantly lower than that of the tumors of the animals treated with control IgG. The immunohistochemical analysis of tumors derived from animals treated with αAM antibody or AM22–52 showed a clear decrease in microvessel density, with an 80% to 90% reduction in endothelial cells within the tumor, which is consistent with the role of AM in endothelial cell survival and recruitment. In agreement with previous studies, these data demonstrate that AM could induce neovascularization and vessel stabilization.
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      • Kaafarani I.
      • Fernandez-Sauze S.
      • Berenguer C.
      • et al.
      Targeting adrenomedullin receptors with systemic delivery of neutralizing antibodies inhibits tumor angiogenesis and suppresses growth of human tumor xenografts in mice.
      • Nikitenko L.L.
      • Fox S.B.
      • Kehoe S.
      • Rees M.C.
      • Bicknell R.
      Adrenomedullin and tumor angiogenesis.
      Recently, we showed that the blockade of AM signaling selectively targets unstable tumor neovessels through rapid disengagement of the VE-cadherin/β-catenin complex, destabilization of the cytoskeleton organization of endothelial cells, and subsequent apoptosis-mediated cell death.
      • Khalfaoui-Bendriss G.
      • Dussault N.
      • Fernandez-Sauze S.
      • et al.
      Adrenomedullin blockade induces regression of tumor neovessels through interference with vascular endothelial-cadherin signalling.
      Increased tumor vessel density appears to negatively affect the survival of patients with MPM,
      • Kumar-Singh S.
      • Vermeulen P.B.
      • Weyler J.
      • et al.
      Evaluation of tumor angiogenesis as a prognostic marker in malignant mesothelioma.
      and antiangiogenic targeted therapies could be useful for MPM treatment.
      • Greillier L.
      • Marco S.
      • Barlesi F.
      Targeted therapies in malignant pleural mesothelioma: a review of clinical studies.
      Of the antiangiogenic therapies, bevacizumab, which is a humanized monoclonal antibody directed against vascular endothelial growth factor, has been assessed in combination with first-line chemotherapy in numerous phase II clinical trials, and the final results of the randomized phase III trial Mesothelioma Avastin Plus Pemetrexin-Cisplatin Study (NCT00651456) have recently demonstrated that the triplet cisplatin–pemetrexed–bevacizumab significantly increases overall survival in patients with MPM.
      • Zalcman G.
      • Margery J.
      • Scherpereel A.
      • et al.
      Bevacizumab 15mg/kg plus cisplatin-pemetrexed (CP) triplet versus CP doublet in malignant pleural mesothelioma (MPM): Results of the IFCT-GFPC-0701 MAPS randomized phase 3 trial.
      Because the vascular endothelial growth factor and AM pathways are complementarily involved in tumor neoangiogenesis,
      • Guidolin D.
      • Albertin G.
      • Spinazzi R.
      • et al.
      Adrenomedullin stimulates angiogenic response in cultured human vascular endothelial cells: involvement of the vascular endothelial growth factor receptor 2.
      targeting both pathways concomitantly or sequentially might optimize the effects of either antiangiogenic strategy in the treatment of MPM.
      Like blood vessel angiogenesis, lymphangiogenesis has attracted attention as an important initial step in tumor pathogenesis.
      • Stacker S.A.
      • Achen M.G.
      • Jussila L.
      • Baldwin M.E.
      • Alitalo K.
      Lymphangiogenesis and cancer metastasis.
      • He Y.
      • Kozaki K.
      • Karpanen T.
      • et al.
      Suppression of tumor lymphangiogenesis and lymph node metastasis by blocking vascular endothelial growth factor receptor 3 signaling.
      Intratumoral or peritumoral lymphangiogenesis increases the risk for metastasis in animal models and in human tumors.
      • Stacker S.A.
      • Achen M.G.
      • Jussila L.
      • Baldwin M.E.
      • Alitalo K.
      Lymphangiogenesis and cancer metastasis.
      To determine whether blockade of AM signaling in MSTO-211H xenografts could impair lymphatic vessel growth, we analyzed the growth of lymphatic vessels within the tumor tissue using a murine LYVE-1 antibody. LYVE-1–positive lymphatic vessels were observed in the control tumors and were significantly decreased in the tumors treated with AM22–52 or αAM antibody. These results suggest that AM signaling induces the growth of lymphatic vessels as a part of its protumorigenic functions. These findings are in agreement with recent data suggesting that tumor-secreted AM is a critical factor in driving tumor lymphangiogenesis.
      • Berenguer-Daize C.
      • Boudouresque F.
      • Bastide C.
      • et al.
      Adrenomedullin blockade suppresses growth of human hormone-independent prostate tumor xenograft in mice.
      • Karpinich N.O.
      • Kechele D.O.
      • Espenschied S.T.
      • Willcockson H.H.
      • Fedoriw Y.
      • Caron K.M.
      Adrenomedullin gene dosage correlates with tumor and lymph node lymphangiogenesis.
      Recently, we demonstrated that activation of AM receptors by AM induces proliferation, migration, invasion, and survival of lymphatic endothelial cells (LECs), thus suggesting an important role for AM in establishing functional lymphatic vessels during tumor growth.
      • Berenguer-Daize C.
      • Boudouresque F.
      • Bastide C.
      • et al.
      Adrenomedullin blockade suppresses growth of human hormone-independent prostate tumor xenograft in mice.
      In this context, the current study provides evidence suggesting that the activation of the AM/AM receptor signaling pathway through upregulated AM expression
      • Fernandez-Sauze S.
      • Delfino C.
      • Mabrouk K.
      • et al.
      Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
      • Berenguer-Daize C.
      • Boudouresque F.
      • Bastide C.
      • et al.
      Adrenomedullin blockade suppresses growth of human hormone-independent prostate tumor xenograft in mice.
      • Berenguer C.
      • Boudouresque F.
      • Dussert C.
      • et al.
      Adrenomedullin, an autocrine/paracrine factor induced by androgen withdrawal, stimulates “neuroendocrine phenotype” in LNCaP prostate tumor cells.
      may be important not only for supporting tumor growth and neoangiogenesis but also for promoting tumor lymphangiogenesis. Thus, targeting the AM system or pharmacologically targeting AM signaling may provide a new avenue for the inhibition of both tumor angiogenesis and lymphangiogenesis.
      In conclusion, our findings suggest that the AM system is a relevant target for treatment of MPM and thus warrants additional preclinical and clinical research into its effects on the development of MPM.

      Acknowledgments

      This study was supported by grants from the Institut National de la Santé et la Recherche Medicale (INSERM), AP-HM, and L'Association pour la Recherche sur les Tumeurs Cérébrales. L'Association pour la Recherche sur les Tumeurs Cérébrales supported Asma Tounsi financially. The authors thank V. Gagna for her excellent secretarial assistance.

      Supplementary Data

      Figure thumbnail figs1
      Figure thumbnail figs2
      Figure thumbnail figs3

      References

        • Carbone M.
        • Kratzke R.A.
        • Testa J.R.
        The pathogenesis of mesothelioma.
        Semin Oncol. 2002; 29: 2-17
        • Ramalingam S.S.
        • Belani C.P.
        Recent advances in the treatment of malignant pleural mesothelioma.
        J Thorac Oncol. 2008; 3: 1056-1064
        • Vogelzang N.J.
        • Rusthoven J.J.
        • Symanowski J.
        • et al.
        Phase III study of pemetrexed in combination with cisplatin versus cisplatin alone in patients with malignant pleural mesothelioma.
        J Clin Oncol. 2003; 21: 2636-2644
        • Folkman J.
        What is the evidence that tumors are angiogenesis dependent?.
        J Natl Cancer Inst. 1990; 82: 4-6
        • Schmitter D.
        • Lauber B.
        • Fagg B.
        • Stahel R.A.
        Hematopoietic growth factors secreted by seven human pleural mesothelioma cell lines: interleukin-6 production as a common feature.
        Int J Cancer. 1992; 51: 296-301
        • Monti G.
        • Jaurand M.C.
        • Monnet I.
        • et al.
        Intrapleural production of interleukin 6 during mesothelioma and its modulation by gamma-interferon treatment.
        Cancer Res. 1994; 54: 4419-4423
        • Antony V.B.
        • Hott J.W.
        • Godbey S.W.
        • Holm K.
        Angiogenesis in mesotheliomas. Role of mesothelial cell derived IL-8.
        Chest. 1996; 109: 21S-22S
        • Marzo A.L.
        • Fitzpatrick D.R.
        • Robinson B.W.
        • Scott B.
        Antisense oligonucleotides specific for transforming growth factor beta2 inhibit the growth of malignant mesothelioma both in vitro and in vivo.
        Cancer Res. 1997; 57: 3200-3207
        • Mutsaers S.E.
        • McAnulty R.J.
        • Laurent G.J.
        • Versnel M.A.
        • Whitaker D.
        • Papadimitriou J.M.
        Cytokine regulation of mesothelial cell proliferation in vitro and in vivo.
        Eur J Cell Biol. 1997; 72: 24-29
        • Kumar-Singh S.
        • Van Marck E.
        Differentiation and differentiating patterns in mesothelioma.
        in: Peters G.A. Peters B. Sourcebook on Asbestos Diseases. Lexus Publishing, Charlottesville, VA1998: 213-284
        • Greig R.
        • Dunnington D.
        • Murthy U.
        • Anzano M.
        Growth factors as novel therapeutic targets in neoplastic disease.
        Cancer Surv. 1988; 7: 653-674
        • Fernandez-Sauze S.
        • Delfino C.
        • Mabrouk K.
        • et al.
        Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors.
        Int J Cancer. 2004; 108: 797-804
        • De Angeli S.
        • Di Liddo R.
        • Buoro S.
        • et al.
        New immortalized human stromal cell lines enhancing in vitro expansion of cord blood hematopoietic stem cells.
        Int J Mol Med. 2004; 13: 363-371
        • Hay D.
        • Walker C.
        • Poyner D.
        Adrenomedullin and calcitonin gene–related peptide receptors in endocrine-related cancers: opportunities and challenges.
        Endocr Relat Cancer. 2011; 18: C1-C14
        • Hinson J.P.
        • Kapas S.
        • Smith D.M.
        Adrenomedullin, a multifunctional regulatory peptide.
        Endocr Rev. 2000; 21: 138-167
        • Zhao Y.
        • Hague S.
        • Manek S.
        • Zhang L.
        • Bicknell R.
        • Rees M.C.
        PCR display identifies tamoxifen induction of the novel angiogenic factor adrenomedullin by a nonestrogenic mechanism in the human endometrium.
        Oncogene. 1998; 16: 409-415
        • McLatchie L.M.
        • Fraser N.J.
        • Main M.J.
        • et al.
        RAMPs regulate the transport and ligand specificity of the calcitonin-receptor-like receptor.
        Nature. 1998; 393: 333-339
        • Poyner D.R.
        • Sexton P.M.
        • Marshall I.
        • et al.
        International Union of Pharmacology. XXXII. The mammalian calcitonin gene–related peptides, adrenomedullin, amylin, and calcitonin receptors.
        Pharmacol Rev. 2002; 54: 233-246
        • Zudaire E.
        • Martinez A.
        • Cuttitta F.
        Adrenomedullin and cancer.
        Regul Pept. 2003; 112: 175-183
        • Deville J.L.
        • Bartoli C.
        • Berenguer C.
        • et al.
        Expression and role of adrenomedullin in renal tumors and value of its mRNA levels as prognostic factor in clear-cell renal carcinoma.
        Int J Cancer. 2009; 125: 2307-2315
        • Kocemba K.
        • van Andel H.
        • de Haan-kramer A.
        • et al.
        The hypoxia target adrenomedullin is aberrantly expressed in multiple myeloma and promotes angiogenesis.
        Leukemia. 2013; 27: 1729-1737
        • Shichiri M.
        • Hirata Y.
        Regulation of cell growth and apoptosis by adrenomedullin.
        Hypertens Res. 2003; 26: S9-S14
        • Belloni A.
        • Treijter M.
        • Malendowicz L.
        • Nussdorf G.
        Adrenomedullin stimulates proliferation and inhibits apoptosis of immature rat thymocytes cultured in vitro.
        Peptides. 2003; 24: 295-300
        • Uzan B.
        • Ea H.
        • Launay J.
        • et al.
        A critical role for adrenomedullin-calcitonin receptor-like receptor in regulating rheumatoid fibroblast-like synoviocyte apoptosis.
        J Immunol. 2006; 176: 5548-5558
        • Martinez A.
        • Vos M.
        • Guedez L.
        • et al.
        The effects of adrenomedullin overexpression in breast tumor cells.
        J Natl Cancer Inst. 2002; 94: 1226-1237
        • Ouafik L.
        • Sauze S.
        • Boudouresque F.
        • et al.
        Neutralization of adrenomedullin inhibits the growth of human glioblastoma cell lines in vitro and suppresses tumor xenograft growth in vivo.
        Am J Pathol. 2002; 160: 1279-1292
        • Kaafarani I.
        • Fernandez-Sauze S.
        • Berenguer C.
        • et al.
        Targeting adrenomedullin receptors with systemic delivery of neutralizing antibodies inhibits tumor angiogenesis and suppresses growth of human tumor xenografts in mice.
        FASEB J. 2009; 23: 3424-3435
        • Berenguer-Daize C.
        • Boudouresque F.
        • Bastide C.
        • et al.
        Adrenomedullin blockade suppresses growth of human hormone-independent prostate tumor xenograft in mice.
        Clin Cancer Res. 2013; 19: 6138-6150
        • Oehler M.K.
        • Hague S.
        • Rees M.C.
        • Bicknell R.
        Adrenomedullin promotes formation of xenografted endometrial tumors by stimulation of autocrine growth and angiogenesis.
        Oncogene. 2002; 21: 2815-2821
        • Karpinich N.O.
        • Kechele D.O.
        • Espenschied S.T.
        • Willcockson H.H.
        • Fedoriw Y.
        • Caron K.M.
        Adrenomedullin gene dosage correlates with tumor and lymph node lymphangiogenesis.
        FASEB J. 2013; 27: 590-600
        • Ishikawa T.
        • Chen J.
        • Wang J.
        • et al.
        Adrenomedullin antagonist suppresses in vivo growth of human pancreatic cancer cells in SCID mice by suppressing angiogenesis.
        Oncogene. 2003; 22: 1238-1242
        • Ramachandran V.
        • Arumugam T.
        • Hwang R.F.
        • Greenson J.K.
        • Simeone D.M.
        • Logsdon C.D.
        Adrenomedullin is expressed in pancreatic cancer and stimulates cell proliferation and invasion in an autocrine manner via the adrenomedullin receptor, ADMR.
        Cancer Res. 2007; 67: 2666-2675
        • Berenguer C.
        • Boudouresque F.
        • Dussert C.
        • et al.
        Adrenomedullin, an autocrine/paracrine factor induced by androgen withdrawal, stimulates “neuroendocrine phenotype” in LNCaP prostate tumor cells.
        Oncogene. 2008; 27: 506-518
        • Albini A.
        • Iwamoto Y.
        • Kleinman H.K.
        • et al.
        A rapid in vitro assay for quantitating the invasive potential of tumor cells.
        Cancer Res. 1987; 47: 3239-3245
        • Ilan N.
        • Mahooti S.
        • Madri J.A.
        Distinct signal transduction pathways are utilized during the tube formation and survival phases of in vitro angiogenesis.
        J Cell Sci. 1998; 111: 3621-3631
        • Yu Y.
        • Sato J.D.
        MAP kinases, phosphatidylinositol 3-kinase, and p70 S6 kinase mediate the mitogenic response of human endothelial cells to vascular endothelial growth factor.
        J Cell Physiol. 1999; 178: 235-246
        • Scherpereel A.
        • Astoul P.
        • Baas P.
        • et al.
        Guidelines of the European Respiratory Society and the European Society of Thoracic Surgeons for the management of malignant pleural mesothelioma.
        Eur Respir J. 2010; 35: 479-495
        • Brown J.M.
        • Giaccia A.J.
        The unique physiology of solid tumors: opportunities (and problems) for cancer therapy.
        Cancer Res. 1998; 58: 1408-1416
        • Metellus P.
        • Voutsinos-Porche B.
        • Nanni-Metellus I.
        • et al.
        Adrenomedullin expression and regulation in human glioblastoma, cultured human glioblastoma cell lines and pilocytic astrocytoma.
        Eur J Cancer. 2011; 47: 1727-1735
        • Kitamura K.
        • Kangawa K.
        • Kawamoto M.
        • et al.
        Adrenomedullin: a novel hypotensive peptide isolated from human pheochromocytoma.
        Biochem Biophys Res Commun. 1993; 192: 553-560
        • Kato H.
        • Shichiri M.
        • Manumo F.
        • et al.
        Adrenomedullin as an autocrine/paracrine apoptosis survival factor for rat endothelial cells.
        Endocrinology. 1997; 138: 2615-2620
        • Nikitenko L.L.
        • Fox S.B.
        • Kehoe S.
        • Rees M.C.
        • Bicknell R.
        Adrenomedullin and tumor angiogenesis.
        Br J Cancer. 2006; 94: 1-7
        • Khalfaoui-Bendriss G.
        • Dussault N.
        • Fernandez-Sauze S.
        • et al.
        Adrenomedullin blockade induces regression of tumor neovessels through interference with vascular endothelial-cadherin signalling.
        Oncotarget. 2015; 6: 7536-7553
        • Kumar-Singh S.
        • Vermeulen P.B.
        • Weyler J.
        • et al.
        Evaluation of tumor angiogenesis as a prognostic marker in malignant mesothelioma.
        J Pathol. 1997; 182: 211-216
        • Greillier L.
        • Marco S.
        • Barlesi F.
        Targeted therapies in malignant pleural mesothelioma: a review of clinical studies.
        Anticancer Drugs. 2011; 22: 199-205
        • Zalcman G.
        • Margery J.
        • Scherpereel A.
        • et al.
        Bevacizumab 15mg/kg plus cisplatin-pemetrexed (CP) triplet versus CP doublet in malignant pleural mesothelioma (MPM): Results of the IFCT-GFPC-0701 MAPS randomized phase 3 trial.
        J Clin Oncol. 2015; 33 (abstract 7500)
        • Guidolin D.
        • Albertin G.
        • Spinazzi R.
        • et al.
        Adrenomedullin stimulates angiogenic response in cultured human vascular endothelial cells: involvement of the vascular endothelial growth factor receptor 2.
        Peptides. 2008; 29: 2013-2023
        • Stacker S.A.
        • Achen M.G.
        • Jussila L.
        • Baldwin M.E.
        • Alitalo K.
        Lymphangiogenesis and cancer metastasis.
        Nat Rev Cancer. 2002; 2: 573-583
        • He Y.
        • Kozaki K.
        • Karpanen T.
        • et al.
        Suppression of tumor lymphangiogenesis and lymph node metastasis by blocking vascular endothelial growth factor receptor 3 signaling.
        J Natl Cancer Inst. 2002; 94: 819-825