Identification of MAGI1 as a tumor-suppressor protein induced by cyclooxygenase-2 inhibitors in colorectal cancer cells.

Cyclooxyganase-2 (COX-2), a rate-limiting enzyme in the prostaglandin synthesis pathway, is overexpressed in many cancers and contributes to cancer progression through tumor cell-autonomous and paracrine effects. Regular use of non-steroidal anti-inflammatory drugs or selective COX-2 inhibitors (COXIBs) reduces the risk of cancer development and progression, in particular of the colon. The COXIB celecoxib is approved for adjunct therapy in patients with Familial adenomatous polyposis at high risk for colorectal cancer (CRC) formation. Long-term use of COXIBs, however, is associated with potentially severe cardiovascular complications, which hampers their broader use as preventive anticancer agents. In an effort to better understand the tumor-suppressive mechanisms of COXIBs, we identified MAGUK with Inverted domain structure-1 (MAGI1), a scaffolding protein implicated in the stabilization of adherens junctions, as a gene upregulated by COXIB in CRC cells and acting as tumor suppressor. Overexpression of MAGI1 in CRC cell lines SW480 and HCT116 induced an epithelial-like morphology; stabilized E-cadherin and β-catenin localization at cell-cell junctions; enhanced actin stress fiber and focal adhesion formation; increased cell adhesion to matrix proteins and suppressed Wnt signaling, anchorage-independent growth, migration and invasion in vitro. Conversely, MAGI1 silencing decreased E-cadherin and β-catenin localization at cell-cell junctions; disrupted actin stress fiber and focal adhesion formation; and enhanced Wnt signaling, anchorage-independent growth, migration and invasion in vitro. MAGI1 overexpression suppressed SW480 and HCT116 subcutaneous primary tumor growth, attenuated primary tumor growth and spontaneous lung metastasis in an orthotopic model of CRC, and decreased the number and size of metastatic nodules in an experimental model of lung metastasis. Collectively, these results identify MAG1 as a COXIB-induced inhibitor of the Wnt/β-catenin signaling pathway, with tumor-suppressive and anti-metastatic activity in experimental colon cancer.

Antibody-conjugated MHC class I tetramers can target tumor cells for specific lysis by T lymphocytes.

To demonstrate that antibody-guided targeting of antigenic MHC class I-peptide tetramer on tumor cells can render them susceptible to lysis by relevant cytotoxic T lymphocytes (CTL), biotinylated HLA-A*0201/Flu matrix peptide complexes were tetramerized on streptavidin molecules previously coupled to Fab' fragments from monoclonal antibodies (mAb) specific for cell surface markers such as carcinoembryonic antigen (CEA), ErbB-2 or CD20. Flow cytometry analysis showed that coating of the HLA-A2-peptide complexes on the four HLA-A2-negative human cancer lines tested (including a CEA-positive colon carcinoma, an ErbB-2(+) breast carcinoma and two CD20(+) B lymphomas) was entirely dependent upon the specificity of the conjugated antibody fragments. More importantly, HLA-A2-restricted Flu matrix peptide-specific CTL were then found to lyse specifically and efficiently the MHC-coated target cells. These results open the way to the development of new immunotherapy strategies based on antibody targeting of MHC class I-peptide complexes.

TAT-RasGAP317-326-mediated tumor cell death sensitization can occur independently of Bax and Bak.

The increase of cancer specificity and efficacy of anti-tumoral agents are prime strategies to overcome the deleterious side effects associated with anti-cancer treatments. We described earlier a cell-permeable protease-resistant peptide derived from the p120 RasGAP protein, called TAT-RasGAP317-326, as being an efficient tumor-specific sensitizer to apoptosis induced by genotoxins in vitro and in vivo. Bcl-2 family members regulate the intrinsic apoptotic response and as such could be targeted by TAT-RasGAP317-326. Our results indicate that the RasGAP-derived peptide increases cisplatin-induced Bax activation. We found no evidence, using in particular knock-out cells, of an involvement of other Bcl-2 family proteins in the tumor-specific sensitization activity of TAT-RasGAP317-326. The absence of Bax and Bak in mouse embryonic fibroblasts rendered them resistant to cisplatin-induced apoptosis and consequently to the sensitizing action of the RasGAP-derived peptide. Surprisingly, in the HCT116 colon carcinoma cell line, the absence of Bax and Bak did not prevent cisplatin-induced apoptosis and the ability of TAT-RasGAP317-326 to augment this response. Our study also revealed that p53, while required for an efficient genotoxin-induced apoptotic response, is dispensable for the ability of the RasGAP-derived peptide to improve the capacity of genotoxins to decrease long-term survival of cancer cells. Hence, even though genotoxin-induced Bax activity can be increased by TAT-RasGAP317-326, the sensitizing activity of the RasGAP-derived peptide can operate in the absence of a functional mitochondrial intrinsic death pathway.

Behavior of endogenous tumor-associated macrophages assessed in vivo using a functionalized nanoparticle.

Tumor-associated macrophages (TAMs) invade the tumor stroma in many cancers, yet their role is incompletely understood. To visualize and better understand these critical cells in tumor progression, we screened a portfolio of rationally selected, injectable agents to image endogenous TAMs ubiquitously in three different cancer models (colon carcinoma, lung adenocarcinoma, and soft tissue sarcoma). AMTA680, a functionally derivatized magneto-fluorescent nanoparticle, labeled a subset of myeloid cells with an "M2" macrophage phenotype, whereas other neighboring cells, including tumor cells and a variety of other leukocytes, remained unlabeled. We further show that AMTA680-labeled endogenous TAMs are not altered and can be tracked noninvasively at different resolutions and using various imaging modalities, e.g., fluorescence molecular tomography, magnetic resonance imaging, and multiphoton and confocal intravital microscopy. Quantitative assessment of TAM distribution and activity in vivo identified that these cells cluster in delimited foci within tumors, show relatively low motility, and extend cytoplasmic protrusions for prolonged physical interactions with neighboring tumor cells. Noninvasive imaging can also be used to monitor TAM-depleting regimen quantitatively. Thus, AMTA680 or related cell-targeting agents represent appropriate injectable vehicles for in vivo analysis of the tumor microenvironment.

Site-specific conjugation of a radioiodinated phenethylamine derivative to a monoclonal antibody results in increased radioactivity localization in tumor.

The preparation of a novel radioiodination reagent, the (aminooxy)acetyl derivative of (p-[125]-iodophenyl)ethylamine, is described. Conventional radioiodination of proteins involves the formation of iodotyrosine residues, but for in vivo applications such as thyroid or stomach immunoscintigraphy, the susceptibility of these residues to tissue dehalogenases constitutes a serious disadvantage. Using our new compound, which has a particularly nonreactive aromatic ring, we confirm and extend studies published by other workers indicating the much greater in vivo stability of iodophenyl compounds compared to the more conventional iodophenolic ones. In addition, the aminooxy group of our reagent gives a stable and specific linkage to aldehyde groups formed by periodate oxidation on the sugar moiety of antibody molecules. In vitro, favorable binding activity and high stability was obtained with a (([125I]iodoaryl)amino)oxy labeled monoclonal antibody directed against carcinoembryonic antigen. In vivo, using paired labeling experiments in nude mice bearing colon carcinoma xenografts, the (([125I]iodoaryl)amino)oxy-MAb (MAb = monoclonal antibody) was compared with the same MAb 131I-labeled by conventional chloramine-T method. Tumor 125I concentration of (arylamino)oxy MAb (measured as percent injected dose per gram) was significantly higher as compared to values obtained with a conventionally labeled 131I antibody. Additionally, thyroid uptake, an indicator of iodine release from the antibody, was up to 25 times lower after injection of 125I-MAb obtained by the new method as compared to the conventionally iodinated 131I-MAb.

Attenuation of colon inflammation through activators of the retinoid X receptor (RXR)/peroxisome proliferator-activated receptor gamma (PPARgamma) heterodimer. A basis for new therapeutic strategies.

The peroxisome proliferator-activated receptor gamma (PPARgamma) is highly expressed in the colon mucosa and its activation has been reported to protect against colitis. We studied the involvement of PPARgamma and its heterodimeric partner, the retinoid X receptor (RXR) in intestinal inflammatory responses. PPARgamma(1/)- and RXRalpha(1/)- mice both displayed a significantly enhanced susceptibility to 2,4,6-trinitrobenzene sulfonic acid (TNBS)-induced colitis compared with their wild-type littermates. A role for the RXR/PPARgamma heterodimer in the protection against colon inflammation was explored by the use of selective RXR and PPARgamma agonists. TNBS-induced colitis was significantly reduced by the administration of both PPARgamma and RXR agonists. This beneficial effect was reflected by increased survival rates, an improvement of macroscopic and histologic scores, a decrease in tumor necrosis factor alpha and interleukin 1beta mRNA levels, a diminished myeloperoxidase concentration, and reduction of nuclear factor kappaB DNA binding activity, c-Jun NH(2)-terminal kinase, and p38 activities in the colon. When coadministered, a significant synergistic effect of PPARgamma and RXR ligands was observed. In combination, these data demonstrate that activation of the RXR/PPARgamma heterodimer protects against colon inflammation and suggest that combination therapy with both RXR and PPARgamma ligands might hold promise in the clinic due to their synergistic effects.

Bcl9/Bcl9l are critical for Wnt-mediated regulation of stem cell traits in colon epithelium and adenocarcinomas.

Canonical Wnt signaling plays a critical role in stem cell maintenance in epithelial homeostasis and carcinogenesis. Here, we show that in the mouse this role is critically mediated by Bcl9/Bcl9l, the mammalian homologues of Legless, which in Drosophila is required for Armadillo/beta-catenin signaling. Conditional ablation of Bcl9/Bcl9l in the intestinal epithelium, where the essential role of Wnt signaling in epithelial homeostasis and stem cell maintenance is well documented, resulted in decreased expression of intestinal stem cell markers and impaired regeneration of ulcerated colon epithelium. Adenocarcinomas with aberrant Wnt signaling arose with similar incidence in wild-type and mutant mice. However, transcriptional profiles were vastly different: Whereas wild-type tumors displayed characteristics of epithelial-mesenchymal transition (EMT) and stem cell-like properties, these properties were largely abrogated in mutant tumors. These findings reveal an essential role for Bcl9/Bcl9l in regulating a subset of Wnt target genes involved in controlling EMT and stem cell-related features and suggest that targeting the Bcl9/Bcl9l arm of Wnt signaling in Wnt-activated cancers might attenuate these traits, which are associated with tumor invasion, metastasis, and resistance to therapy.

Active antiviral T-lymphocyte response can be redirected against tumor cells by antitumor antibody x MHC/viral peptide conjugates.

PURPOSE: To redirect an ongoing antiviral T-cell response against tumor cells in vivo, we evaluated conjugates consisting of antitumor antibody fragments coupled to class I MHC molecules loaded with immunodominant viral peptides. EXPERIMENTAL DESIGN: First, lymphochoriomeningitis virus (LCMV)-infected C57BL/6 mice were s.c. grafted on the right flank with carcinoembryonic antigen (CEA)-transfected MC38 colon carcinoma cells precoated with anti-CEA x H-2D(b)/GP33 LCMV peptide conjugate and on the left flank with the same cells precoated with control anti-CEA F(ab')(2) fragments. Second, influenza virus-infected mice were injected i.v., to induce lung metastases, with HER2-transfected B16F10 cells, coated with either anti-HER2 x H-2D(b)/NP366 influenza peptide conjugates, or anti-HER2 F(ab')(2) fragments alone, or intact anti-HER2 monoclonal antibody. Third, systemic injections of anti-CEA x H-2D(b) conjugates with covalently cross-linked GP33 peptides were tested for the growth inhibition of MC38-CEA(+) cells, s.c. grafted in LCMV-infected mice. RESULTS: In the LCMV-infected mice, five of the six grafts with conjugate-precoated MC38-CEA(+) cells did not develop into tumors, whereas all grafts with F(ab')(2)-precoated MC38-CEA(+) cells did so (P = 0.0022). In influenza virus-infected mice, the group injected with cells precoated with specific conjugate had seven times less lung metastases than control groups (P = 0.0022 and P = 0.013). Most importantly, systemic injection in LCMV-infected mice of anti-CEA x H-2D(b)/cross-linked GP33 conjugates completely abolished tumor growth in four of five mice, whereas the same tumor grew in all five control mice (P = 0.016). CONCLUSION: The results show that a physiologic T-cell antiviral response in immunocompetent mice can be redirected against tumor cells by the use of antitumor antibody x MHC/viral peptide conjugates.

Identification of tumor-associated antigens by large-scale analysis of genes expressed in human colorectal cancer.

Despite the high prevalence of colon cancer in the world and the great interest in targeted anti-cancer therapy, only few tumor-specific gene products have been identified that could serve as targets for the immunological treatment of colorectal cancers. The aim of our study was therefore to identify frequently expressed colon cancer-specific antigens. We performed a large-scale analysis of genes expressed in normal colon and colon cancer tissues isolated from colorectal cancer patients using massively parallel signal sequencing (MPSS). Candidates were additionally subjected to experimental evaluation by semi-quantitative RT-PCR on a cohort of colorectal cancer patients. From a pool of more than 6000 genes identified unambiguously in the analysis, we found 2124 genes that were selectively expressed in colon cancer tissue and 147 genes that were differentially expressed to a significant degree between normal and cancer cells. Differential expression of many genes was confirmed by RT-PCR on a cohort of patients. Despite the fact that deregulated genes were involved in many different cellular pathways, we found that genes expressed in the extracellular space were significantly over-represented in colorectal cancer. Strikingly, we identified a transcript from a chromosome X-linked member of the human endogenous retrovirus (HERV) H family that was frequently and selectively expressed in colon cancer but not in normal tissues. Our data suggest that this sequence should be considered as a target of immunological interventions against colorectal cancer.

Identification of new MHC class I-restricted human tumor antigens for immunotherapy

Summary Cancer is a leading cause of morbidity and mortality in Western countries (as an example, colorectal cancer accounts for about 300'000 new cases and 200'000 deaths each year in Europe and in the USA). Despite that many patients with cancer have complete macroscopic clearance of their disease after resection, radiotherapy and/or chemotherapy, many of these patients develop fatal recurrence. Vaccination with immunogenic peptide tumor antigens has shown encouraging progresses in the last decade; immunotherapy might therefore constitute a fourth therapeutic option in the future. We dissect here and critically evaluate the numerous steps of reverse immunology, a forecast procedure to identify antigenic peptides from the sequence of a gene of interest. Bioinformatic algorithms were applied to mine sequence databases for tumor-specific transcripts. A quality assessment of publicly available sequence databanks allowed defining strengths and weaknesses of bioinformatics-based prediction of colon cancer-specific alternative splicing: new splice variants could be identified, however cancer-restricted expression could not be significantly predicted. Other sources of target transcripts were quantitatively investigated by polymerase chain reactions, as cancer-testis genes or reported overexpressed transcripts. Based on the relative expression of a defined set of housekeeping genes in colon cancer tissues, we characterized a precise procedure for accurate normalization and determined a threshold for the definition of significant overexpression of genes in cancers versus normal tissues. Further steps of reverse immunology were applied on a splice variant of the Melan¬A gene. Since it is known that the C-termini of antigenic peptides are directly produced by the proteasome, longer precursor and overlapping peptides encoded by the target sequence were synthesized chemically and digested in vitro with purified proteasome. The resulting fragments were identified by mass spectroscopy to detect cleavage sites. Using this information and based on the available anchor motifs for defined HLA class I molecules, putative antigenic peptides could be predicted. Their relative affinity for HLA molecules was confirmed experimentally with functional competitive binding assays and they were used to search patients' peripheral blood lymphocytes for the presence of specific cytolytic T lymphocytes (CTL). CTL clones specific for a splice variant of Melan-A could be isolated; although they recognized peptide-pulsed cells, they failed to lyse melanoma cells in functional assays of antigen recognition. In the conclusion, we discuss advantages and bottlenecks of reverse immunology and compare the technical aspects of this approach with the more classical procedure of direct immunology, a technique introduced by Boon and colleagues more than 10 years ago to successfully clone tumor antigens.

PGE2-induced colon cancer growth is mediated by mTORC1.

The inflammatory prostaglandin E2 (PGE2) cytokine plays a key role in the development of colon cancer. Several studies have shown that PGE2 directly induces the growth of colon cancer cells and furthermore promotes tumor angiogenesis by increasing the production of the vascular endothelial growth factor (VEGF). The signaling intermediaries implicated in these processes have however not been fully characterized. In this report, we show that the mechanistic target of rapamycin complex 1 (mTORC1) plays an important role in PGE2-induced colon cancer cell responses. Indeed, stimulation of LS174T cells with PGE2 increased mTORC1 activity as observed by the augmentation of S6 ribosomal protein phosphorylation, a downstream effector of mTORC1. The PGE2 EP4 receptor was responsible for transducing the signal to mTORC1. Moreover, PGE2 increased colon cancer cell proliferation as well as the growth of colon cancer cell colonies grown in matrigel and blocking mTORC1 by rapamycin or ATP-competitive inhibitors of mTOR abrogated these effects. Similarly, the inhibition of mTORC1 by downregulation of its component raptor using RNA interference blocked PGE2-induced LS174T cell growth. Finally, stimulation of LS174T cells with PGE2 increased VEGF production which was also prevented by mTORC1 inhibition. Taken together, these results show that mTORC1 is an important signaling intermediary in PGE2 mediated colon cancer cell growth and VEGF production. They further support a role for mTORC1 in inflammation induced tumor growth.

APRIL, a new ligand of the tumor necrosis factor family, stimulates tumor cell growth.

Members of the tumor necrosis factor (TNF) family induce pleiotropic biological responses, including cell growth, differentiation, and even death. Here we describe a novel member of the TNF family designated APRIL (for a proliferation-inducing ligand). Although transcripts of APRIL are of low abundance in normal tissues, high levels of mRNA are detected in transformed cell lines, and in human cancers of colon, thyroid, and lymphoid tissues in vivo. The addition of recombinant APRIL to various tumor cells stimulates their proliferation. Moreover, APRIL-transfected NIH-3T3 cells show an increased rate of tumor growth in nude mice compared with the parental cell line. These findings suggest that APRIL may be implicated in the regulation of tumor cell growth.

Cytokine targeting in tumors using a bispecific antibody directed against carcinoembryonic antigen and tumor necrosis factor alpha.

The use of tumor necrosis factor alpha (TNFalpha) in cancer therapy is limited by its short circulatory half-life and its severe systemic side effects. To overcome these limitations, we evaluated the capability of a bispecific antibody (BAb) directed against carcinoembryonic antigen (CEA) and human TNFalpha to target this cytokine in tumors. A BAb was constructed by coupling the Fab' fragments from an anti-CEA monoclonal antibody (MAb) to the Fab' fragments from an anti-TNFalpha MAb via a stable thioether linkage. The double specificity of the BAb for CEA and TNFalpha was demonstrated using a BIAcoreTM two-step analysis. The affinity constants of the BAb for CEA immobilized on a sensor chip and for soluble TNFalpha added to the CEA-BAb complex were as high as those of the parental MAbs (1.7 x 10(9) M-1 and 6.6 x 10(8) M-1, respectively). The radiolabeled 125I-labeled BAb retained high immunoreactivity with both CEA and TNFalpha immobilized on a solid phase. In nude mice xenografted with the human colorectal carcinoma T380, the 125I-labeled BAb showed a tumor localization and biodistribution comparable to that of 131I-labeled anti-CEA parental F(ab')2 with 25-30% of the injected dose (ID)/g tumor at 24 h and 20% ID/g tumor at 48 h. To target TNFalpha to the tumor, a two-step i.v. injection protocol was used first, in which a variable dose of 125I-labeled BAb was injected, followed 24 or 48 h later by a constant dose of 131I-labeled TNFalpha (1 microg). Mice pretreated with 3 microg of BAb and sacrificed 2, 4, 6, or 8 h after the injection of TNFalpha showed a 1.5- to 2-fold increased concentration of 131I-labeled TNFalpha in the tumor as compared to control mice, which received TNFalpha alone. With a higher dose of BAb (25 microg), mice showed a better targeting of TNFalpha with a 3.2-fold increased concentration of 131I-labeled TNFalpha in the tumor: 9.3% versus 2.9% ID/g in control mice 6 h after TNFa injection. In a one-step injection protocol using a premixed BAb-TNFalpha preparation, similar results were obtained 6 h postinjection (3.5-fold increased TNFalpha tumor concentration). A longer retention time of TNFalpha was observed leading to an 8.1-fold increased concentration of TNFalpha in the tumor 14 h postinjection (4.4 versus 0.5% ID/g tumor for BAb-treated and control mice, respectively). These results show that our BAb is able, first, to localize in a human colon carcinoma and, there, to immunoabsorb the i.v.-injected TNFalpha, leading to its increased concentration at the tumor site.

Role of transcription factor PROX1 in colon cancer metastasis

Initiation and progression of most colorectal cancers (CRCs) are driven by hyper-activation of the canonical Wnt/ß-catenin/TCF signaling pathway. However, a basal level of activation of this pathway is necessary for intestinal cell homeostasis; thus only CRC-specific effectors of this pathway could be exploited as potential clinical targets. PROX1 is an evolutionary conserved transcription factor with multiple roles in several tissues in embryogenesis, and increasing relevance in cancer. PROX1 is a colon cancer-specific Wnt target in the intestine, thus it might represent a therapeutic target. The role of PROX1 in promoting the transition from early to highly-dysplastic adenoma was previously described [1], Importantly, tumor metastasis is a leading cause of cancer-related mortality. Frequently, micrometastases are already present in patients at the time of diagnosis, therefore better understanding of the mechanisms regulating growth of macrometastatic lesions is important for the development of novel treatment approaches. In this study we showed that PROX1 is expressed in colon cancer stem cell and promotes the outgrowth of metastatic lesions. Firstly, we analyzed the expression of PROX1 in advanced CRCs and their metastases. We found that PROX1 over-expression is a feature of microsatellite stable tumors (~85% of microsatellite stable (MSS) CRCs), which generally have worse prognosis in comparison to microsatellite unstable CRCs. Analysis of primary CRCs and corresponding metastatic lesions showed that PROX1 expression is conserved, or increased in metastases. Further bioinformatics analysis of tumor and metastases gene expression profiles showed that PROX1 is co- expressed with stem cell and progenitor markers. Moreover, in inducible ApcmLgr5-EGFP-lres-CreERT2 model, Prox1+ cells marked a sub-population of Lgr5+ stem cells and subsequent transient amplifying cell population. Orthotopic model of CRC and lung colonization assays in mice demonstrated that PROX1 promotes tumor cell outgrowth in metastatic lesions, while it has no effect on primary tumor growth, invasion, and survival in circulation or cell extravasation. In vitro, PROX1 expressing tumor cells demonstrated strongly increased capacity to form spheroids, and increased survival and proliferation under hypoxic or nutrient-deprivation conditions. By monitoring cellular respiration under these conditions, we found that PROX1 expressing cells exhibit a better metabolic adaptation to changes in fuel source. Autophagy inhibitors, prevented growth both in vitro and in vivo of PROX1 expressing cells. Importantly, conditional inactivation of PROX1 after the establishment of metastases prevented further growth of macroscopic lesions resulting in stable disease. In summary, we identified a novel mechanism underlying the ability of metastatic colon cancer stem and progenitor cells to survive and grow in target organs through metabolic adaptation. Our results establish PROX1 as a key factor of CRC metastatic disease where it promotes survival of metastatic colon cancer stem-like cells, through their metabolic adaptation in sub-optimal microenvironments - L'initiation et la progression de la plupart des cancers colorectaux (CRC) sont entraînées par une hyper-activation de la voie métabolique Wnt/ß- caténine/TCF. Toutefois, un niveau d'activation minimal de Wnt est nécessaire pour l'homéostasie des cellules intestinales ; ainsi seuls des effecteurs spécifiques du CRC- de cette voie pourraient être exploités comme des cibles cliniques potentielles. PROX1 est un facteur de transcription évolutif conservé avec de multiples rôles dans plusieurs tissus durant l'embryogenèse et une pertinence croissante dans le cancer. PROX1 est une cible Wnt spécifique dans le cancer de l'intestin, donc il pourrait représenter une cible thérapeutique. Le rôle de PROX1 durant l'évolution de la maladie d'un stade précoce jusqu'à l'adénome hautement dysplasique a été décrit précédemment. Surtout, la métastase des tumeurs est une cause majeure de mortalité liée au cancer. Souvent, les micro-métastases sont déjà présentes chez les patients au moment du diagnostic, c'est pourquoi une meilleure compréhension des mécanismes régulant la croissance des lésions macrométastatiques est importante pour le développement de nouvelles approches thérapeutiques. Dans cette étude, nous avons prouvé que PROX1 est exprimé dans les cellules souches du cancer du côlon et favorise l'apparition de lésions métastatiques. Nous avons d'abord analysé l'expression de PROX1 dans des CRC avancés ainsi que dans leurs métastases. Nous avons constaté que la surexpression de PROX1 est une caractéristique des tumeurs stables microsatellites (~85% du MSS CRC), qui ont généralement un pronostic défavorable par rapport aux microsatellites CRC instables. L'analyse des CRC primaires et de leurs métastases liées a montré que l'expression de PROX1 est conservée, voire augmentée dans les métastases. A l'aide d'une base de données de tumeurs et métastases, nous avons observé une co- régulation de PROX1 entre cellules souches et marqueurs de progéniteurs mais pas avec des cellules différenciées. De plus, en utilisant un modèle Apcm Lgr5-EGFP-IRES-CreERT2 inductible, les cellules Prox1+ ont marqué une sous-population de cellules LGR& capable de produire une lignée. Un modèle orthotopique de cancer colorectal et des essais de colonisation du poumon chez la souris ont démontré que PROX1 favorise l'excroissance des cellules tumorales dans les lésions métastatiques, alors qu'il n'a aucun effet sur la croissance tumorale primaire, l'invasion ou une extravasation des cellules. In vitro, les cellules tumorales exprimant PROX1 ont démontré une forte augmentation de leur capacité à former des sphéroïdes, ainsi qu'une augmentation de la survie et de la prolifération dans des conditions hypoxiques ou lors de privation de nutriments. En contrôlant la respiration cellulaire dans ces conditions, nous avons constaté que les cellules exprimant PROX1 présentent une meilleure adaptation métabolique à l'évolution des sources de carburant. Des inhibiteurs de l'autophagie, suggérant une approche thérapeutique potentielle, ont tué à la fois in vitro et in vivo les cellules exprimant PROX1. Surtout, l'inactivation conditionnelle de PROX1 après l'apparition de métastases a empêché la croissance des lésions macroscopiques résultant en une maladie stable. En résumé, nous avons identifié un nouveau mécanisme mettant en évidence la capacité des cellules souches du cancer du côlon métastatique à survivre et à se développer dans les organes cibles grâce à l'adaptation métabolique. Nos résultats définissent PROX1 comme un facteur clé du cancer colorectal métastatique en favorisant la survie des cellules souches métastatiques apparentées au cancer du colon grâce à leur adaptation métabolique aux microenvironnements défavorables.

Tissue biomarker development in a multicentre trial context: a feasibility study on the PETACC3 stage II and III colon cancer adjuvant treatment trial.

PURPOSE: We evaluated the feasibility of biomarker development in the context of multicenter clinical trials. EXPERIMENTAL DESIGN: Formalin-fixed, paraffin-embedded (FFPE) tissue samples were collected from a prospective adjuvant colon cancer trial (PETACC3). DNA was isolated from tumor as well as normal tissue and used for analysis of microsatellite instability, KRAS and BRAF genotyping, UGT1A1 genotyping, and loss of heterozygosity of 18 q loci. Immunohistochemistry was used to test expression of TERT, SMAD4, p53, and TYMS. Messenger RNA was retrieved and tested for use in expression profiling experiments. RESULTS: Of the 3,278 patients entered in the study, FFPE blocks were obtained from 1,564 patients coming from 368 different centers in 31 countries. In over 95% of the samples, genomic DNA tests yielded a reliable result. Of the immmunohistochemical tests, p53 and SMAD4 staining did best with reliable results in over 85% of the cases. TERT was the most problematic test with 46% of failures, mostly due to insufficient tissue processing quality. Good quality mRNA was obtained, usable in expression profiling experiments. CONCLUSIONS: Prospective clinical trials can be used as framework for biomarker development using routinely processed FFPE tissues. Our results support the notion that as a rule, translational studies based on FFPE should be included in prospective clinical trials.