Role of MIF/CD74 signaling pathway in the development of pleural mesothelioma

Macrophage migration inhibitory factor (MIF) is a pro-inflammatory cytokine implicated in acute and chronic inflammatory diseases. MIF is overexpressed in various tumors. It displays a number of functions that provide a direct link between the process of inflammation and tumor growth. Our group recently identified the MIF-receptor CD74 as an independent prognostic factor for overall survival in patients with malignant pleural mesothelioma. In the present study, we compared the levels of expression of MIF and CD74 in different human mesothelioma cell lines and investigated their physiopathological functions in vitro and in vivo. Human mesothelioma cells expressed more CD74 and secreted less MIF than non tumoral MeT5A cells, suggesting a higher sensitivity to MIF. In mesothelioma cells, high MIF levels were associated with a high multiplication rate of cells. In vitro, reduction of MIF or CD74 levels in both mesothelioma cell lines showed that the MIF/CD74 signaling pathway promoted tumor cell proliferation and protected MPM cells from apoptosis. Finally, mesothelioma cell lines expressing high CD74 levels had a low tumorigenic potential after xenogeneic implantation in athymic nude mice. All these data highlight the complexity of the MIF/CD74 signaling pathway in the development of mesothelioma.


IntroductIon
Malignant pleural mesothelioma (MPM) is an aggressive cancer of the inner lining of the chest cavity which develops mainly after inhaling asbestos fibers. According to the WHO classification, malignant mesothelioma is either classified as epithelioid (mostly composed of epithelial-shaped cells), sarcomatoid (mostly composed of spindle-shaped cells), or biphasic (composed of both types of cells) [1]. In 2007, the WHO estimated that about 125 million people around the world were exposed to asbestos at work, and that at least 90,000 people died each year from asbestos-related diseases. In the next decades, MPM incidence will continue to increase, even in regions where the commercial use of asbestos has been banned (Europe, Australia and Japan) and will contribute to cancer mortality in countries lacking working protection and/or persisting with its use (Asia and India). While surgery is a valid option for patients with early stage MPM, most patients with locally advanced invasive disease are not amenable to surgical resection [2] and treatment is palliative chemotherapy combining cisplatin and pemetrexed. While this treatment may relieved symptoms, it provides only modest survivals, since median survival averages only 9-18 months from the time of diagnosis.
The exact mechanisms involved in the neoplastic transformation of mesothelial cells are still unknown.
Identifying key genes related to the underlying oncogenic processes and understanding the malignant mesothelial cell regulation pathways are essential to develop more effective treatments. Initial tumor growth depends on increased cell proliferation and reduced cell death, both of which are stimulated by inflammation-driven mechanisms. Tumor cells trigger an intrinsic inflammatory response that builds up a protumorigenic microenvironment [3]. No study has precisely characterized the tumor inflammatory environment in the early development and growth of mesothelioma. Macrophage migration inhibitory factor (MIF) is a pro-inflammatory cytokine involved in both innate and adaptive immunity. Originally described as a T cell-derived product, MIF is released by numerous cell types and is involved in many inflammatory and autoimmune diseases [4]. Furthermore, MIF may be involved in cell proliferation and differentiation and several studies have reported increased MIF mRNA levels in tumor cells and pre-tumor states in prostate [5], colon [6][7][8], and hepatocellular cancers [9], adenocarcinomas of the lung [10], glioblastomas [11,12] and melanomas [13]. Several groups have shown a correlation between MIF expression and cancer prognosis in hepatocellular carcinomas, colon and prostate cancers [8,14,15]. The invariant chain or CD74 was the first MIF surface receptor described [3]. Chemokine receptors CXCR2 and CXCR4 have also been shown to be MIF receptors [16][17][18]. In murine models of human colorectal adenoma [16] and metastatic breast cancer [19], inhibition of MIF expression by genetic deletion or RNA interference decreased tumor progression and metastasis. A reduction of tumor growth was also observed in sub-cutaneous human neuroblastoma [20] and prostate cancer xenografts [21], after inhibition of MIF expression by MIF antisens transfection or RNA interference or after treatment with anti-MIF antibodies or MIF inhibitors. Such a decrease of MIF activity was associated with inhibition of tumor angiogenesis [21].
Our group has recently shown that the majority of malignant mesothelial tumor cells express MIF and its receptor CD74, with a homogenous distribution between the different histological subtypes [22]. We also demonstrated that high levels of CD74 were an independent prognostic factor for prolonged overall survival in MPM patients. To clarify the potential involvement of MIF in MPM, we investigated the expression and the role of MIF and its receptors (CD74, CXCR2 and CXCR4) in viability, proliferation and in vivo tumorigenesis of human MPM cells.

results
In order to identify the role of MIF in MPM, mRNA and protein levels of MIF-receptors CD74 (Figure 1 (Table 1).

MPM cell lines expressed higher levels of cd74 compared to mesothelial cell line Met5A
Expression of CD74 in MeT5A cells was weak to absent ( Figure 1A and 1B). CD74 mRNA expression levels of all MPM cell lines studied were significantly higher than that of MeT5A cells ( Figure 1A, n = 3; P < 0.05). CD74 total protein expression in MPM cell lines was higher than that of MeT5A cells except for JL-1 and MSTO cell lines ( Figure 1B). Cell surface expression of CD74 was not detected using flow cytometry in all MPM cell lines studied and MeT5A cells (Supplementary Figure S1). Previous studies about cell surface CD74 showed that surface expression of newly synthesized CD74 complexes concern only few percents of cellular CD74 and is followed by a rapid internalization to the endosomal pathway [27] complicating cell surface detection of these complexes.
Thus, CD74 appeared to be primarily expressed in malignant mesothelial cells, indicating that such tumor cells may be prone to stimulation with MIF.

MPM cell lines expressed similar levels of cXcr4 than mesothelial cell line Met5A
CXCR4 mRNA and protein levels (Figure 2A and 2B) were assessed by RT-qPCR and western blotting. No significant difference in CXCR4 expression levels between MPM cell lines compared to MeT5A was observed. A difference in cell surface expression of CXCR4 (P < 0.05) was detected between MeT5A and the MPM JL1 cells ( Figure 3C) with a median fluorescence intensity of 2.9 ± 0.3 (n = 7) and 5.4 ± 1.7 (n = 8) for MeT5A and JL1 respectively. In MPM cell lines, CXCR4 expression levels varied between the different cell lines studied and these levels of expression were not related to the histological type of the MPM.
The chemokine receptor CXCR2 has also been described as a receptor for MIF. CXCR2 mRNA levels were weak to absent in the MeT5A cells as well as MPM cells (data not shown) suggesting a very poor protein expression of CXCR2 in MPM and mesothelial cells.

MPM cell lines secreted lower levels of MIF compared to mesothelial cell line Met5A
MIF mRNA and protein levels ( Figure 3A and 3B) were assessed by RT-qPCR and western blotting. No significant difference in MIF expression levels between MPM cell lines compared to MeT5A was observed. In MPM cell lines, MIF expression levels varied between the different cell lines studied and these levels of expression were not related to the histological type of the MPM.
During tumorigenesis, tumor cells secrete growth factors and cytokines (including MIF) which amplify tumor cell transformation, activate their proliferation and modify the activation state of surrounding inflammatory cells. In order to determine whether MeT5A and MPM cells spontaneously secrete MIF in the culture medium, secreted MIF levels by MeT5A and MPM cell lines were measured after 48 h-culture by ELISA ( Figure 3C). In all MPM cell lines studied, accumulated MIF concentrations in the supernatant reached levels significantly lower (from 2-fold to 14-fold lower for H28 and JL-1 respectively, n = 5; P < 0.05) than the MIF secretion rate of MeT5A.
In summary, most MPM cell lines expressed higher levels of CD74 and secreted lower levels of MIF than mesothelial cell line MeT5A suggesting a higher sensitivity to MIF.

MPM cells expressing high levels of MIF/cd74 and secreting high levels of MIF showed high multiplication rate
We compared physiological characteristics, such as cell multiplication, cell proliferation and apoptosis of MeT5A cells with JL-1 and H28 MPM cell lines chosen in regard to their MIF/CD74 expressions and MIF secreting levels. As shown previously, JL-1 cells expressed the lowest MIF/CD74 levels and secreted the lowest MIF level and H28 cells expressed the highest MIF/CD74 levels and secreted the highest MIF level (Figures 1 and 3) of the six MPM cell lines studied.
Finally, chemosensitivity of the three cell lines to cisplatin, a chemotherapeutic drug used in human MPM treatment was assessed. Cell multiplication was evaluated after 24 h of treatment with cisplatin using a MTT assay ( Figure 4D). H28 and MeT5A cells were less sensitive to cisplatin treatment than JL-1 cells as after 24 h of cisplatin treatment, 67.9 ± 33.4% (n = 12) of H28 and 71.8 ± 20.1% (n = 9) of MeT5A cells and only 38.1 ± 13.5% (n = 9) of JL-1 cells recovered compared to untreated cells.
All these data were analysed in regard to MIF-secreted level and CD74 expression in JL-1, H28 and MeT5A cell lines (Figures 1 and 3). Human MPM cell Relative mRNA (A) or protein (B) expression levels were measured by qPCR and western blotting respectively. Data represent the mean values (± SD) of 3 to 6 independent experiments. significant difference Kruskal-Wallis test between normal mesothelial cell line MeT5A and mesothelioma cell lines: *P < 0.05. line H28 (expressing high level of CD74 and secreting high level of MIF) showed a high level of multiplication and proliferation rate associated with a low level of apoptosis rate. In contrast, the MPM cell line JL-1 (expressing low level of CD74 and secreting low level of MIF) showed a low level of multiplication and proliferation rate associated with a high level of apoptosis rate. These data suggested that MIF increased cell viability through increase of proliferation and decrease of apoptosis of MPM cells. Additionally, high level of secreted MIF and CD74 expression was also associated to a lower sensitivity to cisplatin.
In summary, in human MPM H28 and H2052 cell lines, a MIF expression and secretion reduction led to a decrease in cell multiplication due to an increase of cell apoptosis. Additionally, a reduction in CD74 expression led to a decrease of cell multiplication, due to a decrease of cell proliferation and an increase of cell apoptosis.

MPM cells expressing higher level of cd74 showed lower tumorigenic potential
In order to analyse the association between MIF/ CD74 stimulated pathway and the tumorigenic potential of MPM cells, we followed xenografts development of two human MPM cell lines expressing high levels of CD74 (H28 and H2052, see Figure 1B), and two human MPM cell lines expressing low levels of CD74 (JL-1 and MSTO, see Figure 1B). Cell lines were injected individually subcutaneously (s.c.) and intra-pleuraly (i.pl.) into athymic nude mice (immunodeficiency affecting T cells), and tumor development were assessed using caliper (for s.c. tumors) and positron emission tomography/computed tomography (PET/CT) scans (for pleural tumors).
MSTO, JL-1 and H2052 pleural tumors were identified in the thoracic cavity using 2-deoxy-2-[ 18 F] fluoro-D-glucose ([ 18 F]FDG)-PET/CT analyses ( Figure  6B, top panels) in nearly all injected mice (9/10, 5/10 and 5/6, respectively). Post-mortem examination showed that athymic nude mice injected with MSTO, JL-1 and H2052 cells had extensive pleural tumors. Most of tumors were free in the thoracic cavity; some tumors were attached to the thoracic muscles and the diaphragm and few to the lung. No mice injected i.pl. with H28 cells developed tumors until 155 days after injection. In order to estimate the MIF and CD74 expression in intra-pleural MPM tumors developed into athymic nude mice, we performed immunohistochemical staining for MIF, CD74 and the coreceptor of CD74, CD44 on intra-pleural JL-1, H2052 and MSTO tumors. As shown in Figure 7, cytoplasmic MIF was detected in tumor cells of JL-1, H2052 and MSTO tumors. As expected neither CD74 nor its co-receptor CD44 were detected in JL-1 tumor cells. Cytoplasmic and membranous expression of CD74 was detected in H2052 tumor cells. These cells also expressed high amount of cytoplasmic and membranous CD44. Cytoplasmic and membranous CD74 was also detected in MSTO tumor cells as well as CD44. CD44 labeling of MSTO tumors was less intense than that of H2052.
Our in vivo data suggested that CD74/CD44 expression was negatively associated with in vivo tumorigenesis in MPM.

dIscussIon
MIF is a pro-inflammatory cytokine overexpressed in various cancers [5-10, 12, 13]. It may play an important role in carcinogenesis by promoting cell proliferation, tumor angiogenesis and metastasis [26]. Up to now, no study has addressed the role of MIF in MPM. Recently, we have shown that the majority of the malignant mesothelial tumor cells expressed MIF and its receptor CD74, with a homogenous distribution between the different histotypes. We also demonstrated that high expression of CD74 was an independent prognostic factor for prolonged overall survival in MPM patients [22].
In this study, we characterized the expression and impact of MIF on human MPM cell lines. MIF protein expression was confirmed in immortalized mesothelial MeT5A cell line and in several MPM cell lines derived from epithelioid, sarcomatoid and biphasic human MPM. MIF protein expression level was not significantly different between MeT5A and MPM cells, suggesting that de novo synthesis and intracellular protein storage did not differ between normal and malignant mesothelial cells. In contrast, MIF secretion was lower in MPM cells compared to MeT5A cells. These lower levels of secreted MIF could be related to the higher level of expression of MIF-receptor CD74 in MPM cells, conferring a higher sensitivity to MIF of MPM cells compared to normal mesothelial cells. The very low level of CD74 in normal mesothelial cell line MeT5A suggest that these cells are and MeT5A cells were cultured for 24 h. Cell multiplication (A) was measured using a MTT assay. Cell proliferation (b) was estimated by EdU incorporation after 2 h cultured with EdU. Proliferation rate was calculated as the percentage of EdU + -Hoechst + double-positive cells (Zeiss Apotome, 20× magnification, Axiovision 4.6). Cell apoptosis (c) was analyzed using Annexin-V-PE/7-AAD double staining. Effect of cisplatin on cell multiplication (d) was assessed using the MTT assay performed on mesothelioma cells cultured for 24 h with 100 mmol/L of cisplatin. Control was cells cultured for 24 h without cisplatin. Bars are mean values (± SD) for n = 4-15 independent experiments. Kruskal-Wallis test between normal mesothelial cell line MeT5A and mesothelioma cell lines: *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. not sensitive to MIF, explaining in part why their high MIF secretion level was not correlated to their multiplication rate. Cell surface CD74 could not be detected on MPM cells using flow cytometry. Previous studies about cell surface CD74 showed that surface expression of newly synthesized CD74 complexes concern only few percents of cellular CD74 and is followed by a rapid internalization to the endosomal pathway [27] complicating cell surface detection of these complexes. Expression of the chemokine receptors CXCR2 and CXCR4, described as MIF receptors, was characterized as the same levels in mesothelial and MPM cell lines suggesting that these receptors are not implicated in MPM cell multiplication.
In the MPM cell lines studied, we observed that the higher CD74 expression and MIF secreted levels, the higher the multiplication rate was, suggesting that MIF signaling through CD74 could play a role in MPM cell proliferation and apoptosis. This role was confirmed by reducing MIF and CD74 expression in MPM H28 and H2052 cells using specific siRNA. A reduction in CD74 expression led to a decrease in the multiplication rate due to a decrease of the proliferation and an increase of the apoptosis rates. Reducing MIF expression and secretion had the same effect on multiplication suggesting that MIF signal transduction is initiated by binding to CD74. MIF reduction did not change the proliferation rate. This could potentially be explained either by a resting MIF level sufficient to activate proliferation of MPM cells or by pro-stimulatory effect of another CD74 ligand such as the functional homolog of MIF, the D-dopachrome tautomerase [28].
Our data suggest that secreted MIF (or MIF homolog) from tumor cells has pro-stimulatory and anti-apoptotic effects on MPM cells. This pro-stimulatory effect could also be due, in situ, to MIF secreted by cells from the tumor microenvironment, such as tumor-associated  macrophages, tumor-infiltrating lymphocytes, or cancerassociated fibroblasts [29][30][31][32][33]. In several tumoral cells (breast, colon, prostate, lung, …), MIF signaling in cancer is well described to be triggered by its receptor CD74 [21,34,35]. Previously, in tumoral cells of human MPM explants, we showed [22] a co-expression of MIF and CD74 suggesting that MIF effect on MPM cells was triggered in part by binding to CD74. In summary, in MPM cell lines, activated MIF/CD74 pathway has a protumorigenic function by increasing tumor cell proliferation and protecting them from apoptosis.
In vivo data obtained from the development of MPM into athymic nude mice showed at contrary that the lower CD74 expression level is, the higher the tumorigenic potential is. These results should be related to our previous results obtained on human tissue array [22] showing that low CD74 expression level in MPM cells was associated with a low patient survival rate. We also observed an up-regulation of CD74 expression in situ in MSTO-211H tumors. These data therefore showed additional regulatory pathway on CD74 expression in vivo. These different results highlight the complex role of MIF and its receptor CD74 in tumor growth. The effect of MIF and CD74 on carcinogenesis seems to change with the cell type as well as the stage of the cancer. This may be a consequence of the activation of different signaling pathways. Several studies reported the activation of the extracellular signal regulated kinase (ERK)1 and 2 in the mitogen-activated protein kinase pathway, and the PI3K/Akt/SRC signal transduction cascade [3,36,37] subsequently to the binding of MIF to its receptor complex CD74/CD44. These activated pathways upregulate cell proliferation, decrease cell apoptosis and enhance cell migration [21,38]. In contrast, other studies showed that MIF can activate the AMPK pathway, leading in some cancers to a decrease in cell proliferation, cell viability and in their metastatic potential [39,40]. Finally, we observed differences in tumor development after s.c. or i.pl. injection with H2052 cells. H2052 tumors developed slowly but extensively in the thoracic cavity while s.c tumors poorly grew. These differences could be due to differences in the vascularization, growth factors and immune cells in the environment of these both sites. While s.c graft site has the advantage to be easily accessible, it is not representative of the environment in which the tumor originated. Wang Y et al. [41] showed that the histology of human prostate tumors implanted into immunodefficient mice was best in the orthotopically grafted samples and that s.c. grafted tissues had the poorest profile of histopathologic differentiation. The orthotopic graft site provides a tumor microenvironment that closely reflects the clinical situation. The use of orthotopic sites versus other sites may well be important in the metastatic spread patterns of any advance malignant graft. The use of the orthotopic thoracic site for xenografting has not been widespread, due largely to the technical difficulties in reaching, and monitoring tumor development in this location. In our study, the progression of mesothelioma in the pleural cavity was easily monitored by using PET/CT imaging, offering a new valuable tool for evaluation of anticancer therapy for MPM. In the future, in order to clarify the complex role of MIF and/or CD74 from tumor and stromal cells in MPM development, we plan to perform i.pl. syngeneic implantations of murine mesothelioma cells deleted or not in MIF or CD74 expression into wild-type or MIF-deficient or CD74-deficient mice and study tumor development, tumor angiogenesis and identify the inflammatory cells recruited by the host.
In conclusion, the results of the present study showed that most human MPM cell lines expressed higher level of CD74 and secreted lower level of MIF than MeT5A mesothelial cell line. Decreasing MIF or CD74 expression in H28 and H2052 MPM cells reduced multiplication rate of the tumor cells due to a reduction in proliferation and an increase in apoptosis. Finally, in vivo data following xenogeneic human MPM cells implanted into athymic mice suggested that CD74 expression was negatively associated with in vivo tumorigenesis in MPM. All these data showed the complex role of MIF/CD74 pathway on MPM development with, on one hand, a promoting effect on tumor cell viability and on the other hand, a promoting effect on mesothelioma cell-stroma interactions.

total rnA isolation and real-time rt-Pcr
The expression of MIF, CD74, CXCR2, CXCR4, GAPDH, GUSB, EEFLA1 and TBP mRNAs was evaluated by quantitative RT-PCR analysis. Total mRNAs from each cell lines was extracted by Qiagen RNEasy Midi extraction Kit (Qiagen) according to manufacturer's instructions.
Reverse transcription was performed from 0.5 µg of total RNA using PrimerScript reverse transcriptase enzyme (Takara bio inc. Kit) and a mix of random hexamers -oligo d(T) primers, following suppliers instructions.
Real-time RT-PCR was performed on each sample in triplicate performed from 500 ng cDNA using an ABI 7900HT SDS system with Power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, California). SYBR green primers were designed using the program Primer Express v 2.0 (Applied Biosystems) with default parameters and obtained from Invitrogen. Primer sequences for the targeted human genes are available upon request. Results were normalized to the expression levels of GAPDH, GUSB, EEFLA1 and TBP expression genes, used as housekeeping genes. Normalisation factor and fold changes were calculated using the GeNorm method [43].

Biochemical analyses
Thirty thousand cells were seeded onto 24-well microplates in 500 µl/well of complete RPMI medium. Supernatants were removed 48 h after culture, centrifuged 20 min at 1,000 g and stored at −20°C. MIF levels in medium supernatant were detected using the Human MIF Duoset kit (R & D System).

cell proliferation
Cell proliferation was measured by performing 5-ethynyl-20-deoxyuridine (EdU) incorporation assay, using the Click-iT EdU imaging kit (Life Technologies) according to the manufacturer's instructions. Cells were cultured in 8-chamber slides (Lab-Tek permanox chamber slide from Nunc) at a density of 25 × 10 3 cells/cm². After overnight incubation, and serum-starvation for 24 h in RPMI medium supplemented with 1% (w/v) HSA, medium was replaced by complete RPMI, and cells were incubated for another 24 h at 37°C. Then, 5 µM of EdU was added to each chamber and cells were cultured for additional 2 h at 37°C. The cells were fixed with 10% (v/v) formalin for 20 min (RT) and permeabilized with 0.5% (v/v) Triton X-100 for 10 min (RT). After washing with PBS, cells were incubated with the EdU reagent cocktail for 30 min at RT. Then, the cells were stained with mounting medium ProTaqs Mount Fluor (BIOCYC) containing Hoechst33342. The EdU (green) and Hoechst (blue) positive nuclei were counted and the EdU incorporation rate was expressed as the ratio of EdU positive cells to total Hoechst positive cells.

Cell apoptosis analysis
Cell apoptosis was evaluated using Annexin-Vphycoerythrin/7-amino-actinomycin D (Annexin-V-PE/7-AAD) double staining. Cells (2 × 10 5 to 4 × 10 5 ) were cultured for 24 h in non-adherent 35 mm-dishes, in complete RPMI. After washes with cold PBS and Annexin V Binding Buffer (100 mmol/l Hepes, 1.4 mol/l NaCl, 25 mmol/l CaCl 2 ), cells were resuspended in 100 µl of Annexin V Binding Buffer and incubated with 2.5 µl of PE Annexin V (BioLegend #640912) and 7 µl of 7-AAD (BD Pharmingen # 51-2359KC) for 15 min at RT in the dark. Annexin V Binding Buffer (100 µl) were added to each tube and annexin and 7-AAD fluorescences were analysed with an Accuri C6 flow cytometer (BD). For each measurement, 5 × 10 4 cells were counted. Dot plots and histograms were analysed by BD Accuri TM C6 software (BD). Annexin V-positive cells were considered in the early stage of apoptosis; Annexin V-and 7-AADpositive cells were considered in the late stage of apoptosis or necrotic. Annexin V-and 7-AAD-negative cells correspond to the viable cell fraction.
H28 cell apoptosis after transfection with siRNA was evaluated using a terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay. Forty-eight hours after transfection, H28 cells were fixed in 10% (v/v) formalin and TUNEL assay (In Situ Cell Death Detection Kit, TMR red, Roche Diagnostics) was performed according to the manufacturer's instructions. Briefly, cells were permeabilized in 0.1% (v/v) triton-X100 in PBS for 10 min and incubated with the reaction mixture containing the enzyme terminal deoxynucleotidyl transferase and the fluorescent TMR-conjugated dUTP for 1 h at 37°C. Labeled DNA was visualized by fluorescence microscopy.

sub-cutaneous and i.pl. implantations of MPM cells in nude mice
Mice were anaesthetized with isoflurane and MPM cells (MSTO, JL-1, H2052 and H28) were injected s.c.
(1 × 10 6 tumor cells suspended in 200 µl of of a matrix containing 30% of RPMI and 70% of HyStem-C hydrogel (ESI BIO, Alameda, CA, USA)) on the right dorsa and into the left pleural cavity (1 × 10 6 tumor cells suspended in 50 µl of RPMI) of 8-week-old athymic female nude mice nu/nu (Harlan) (n = 5 or 6 per group). Once a week, volumes of s.c. tumors were measured using calipers and calculated using the following formula: (Long axis (mm) x short axis (mm) x short axis (mm))/2 Intra pleural tumors were followed by 2-deoxy-2-[ 18 F]fluoro-D-glucose ([ 18 F]FDG)-PET/computed tomography (CT) scans. PET/CT was performed using a Triumph PET/SPECT/CT system (Trifoil, Chatsworth, CA, USA). The mice were fasted for 12 h and blood glucose was measured before each scan. Mice were anesthetized with 2% isoflurane and were i.v. injected retro-orbitally with 5-6 MBq of [ 18 F]FDG. Mice were then left awake at RT during an uptake time of 60 min. 10 min prior to PET scan, mice were injected intraperitoneally with 700 µL of 132 mg/ml meglumine ioxitalamate (Telebrix, 6% m/v iodide, Guerbet AG, Zürich, Switzerland) to delineate the abdominal region and subjected to CT scans. Images were obtained at 80 kVp, 160 µA, and 1024 projections were acquired during the 360° rotation with a field of view of 71.3 mm (1.7 × magnification). After 60 min of [ 18 F]FDG uptake, PET scans were started for a duration of 20 min. PET scans were reconstructed with the built-in LabPET software using an OSEM3D (20 iterations) algorithm and images were calibrated in Bq/ml by scanning a phantom cylinder. The Triumph XO software, which uses a back-projection engine, was used to reconstruct the CT scans with a matrix of 512 and a voxel size of 0.135 mm. CT scans were co-registered with the PET scans using the plugin Vivid (Trifoil) for Amira (FEI, Hillsboro, OR, USA) and exported as dicom files. The software Osirix (Pixmeo, Bernex, Switzerland) was used to quantitatively analyse the datasets and generate pictures.
At the end of the experiment, all mice were euthanized and closely examined for the presence of thoracic tumors. This study was conducted under protocols revised and approved by the institutional animal care and use committee and by Geneva's veterinarian state office.

statistics
Results were presented as means ± SD. Statistical differences among three or more groups were examined by a Kruskal-Wallis test. Differences between pairs of groups were examined for statistical significance using the unpaired Mann-Whitney U test. A P value < 0.05 was considered as statistically significant.