Mab16d10 Induces Cell Death at Tumoral Pancreatic Cell Surface By

Targeting a Novel Onco-glycoprotein Antigen at Tumoral Pancreatic Cell Surface by mAb16D10 Induces Cell Death

This information is current as Lydie Crescence, Evelyne Beraud, Veronique Sbarra, of September 28, 2021. Jean-Paul Bernard, Dominique Lombardo and Eric Mas J Immunol 2012; 189:3386-3396; Prepublished online 5 September 2012; doi: 10.4049/jimmunol.1102647

http://www.jimmunol.org/content/189/7/3386 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2012/09/05/jimmunol.110264 Material 7.DC2 http://www.jimmunol.org/ References This article cites 57 articles, 20 of which you can access for free at: http://www.jimmunol.org/content/189/7/3386.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 28, 2021 • No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Targeting a Novel Onco-glycoprotein Antigen at Tumoral Pancreatic Cell Surface by mAb16D10 Induces Cell Death

Lydie Crescence,*,† Evelyne Beraud,*,† Veronique Sbarra,*,† Jean-Paul Bernard,*,†,‡ Dominique Lombardo,*,†,‡ and Eric Mas*,†,‡

The mAb16D10 was raised against a pathological onco-glycoform of bile salt-dependent lipase isolated from the pancreatic juice of a patient suffering from a pancreatic adenocarcinoma. We previously showed that mAb16D10 speciﬁcally discriminates human pancreatic tumor tissues from other cancer and nontumor tissues. In this study, we report that mAb16D10 inhibited the proliferation of only human pancreatic tumor cells expressing 16D10 plasma membrane Ag. Interaction of mAb16D10 with its cognate surface Ag on pancreatic cells promoted cell death by activation of the p53- and caspase-dependent apoptotic pathway, and silencing of p53 decreased cell death. The decreased proliferation was also partly due to cell cycle arrest in G1/S phase, mAb16D10 triggering of glycogen synthase kinase-3b (GSK-3b) activation, degradation of b-catenin, and decreased expression of cyclin D1. GSK-3b Downloaded from positively affected p53 expression in pancreatic tumor cells after mAb16D10 binding. Inhibition of GSK-3b activity reversed the effects induced by mAb16D10 in SOJ-6 cells, supporting the pivotal role of GSK-3b signaling in the mechanisms of action induced by mAb16D10. Also, mAb16D10 cell treatment led to membrane overexpression of E-cadherin. Both E-cadherin and tumor Ag were localized in membrane lipid cholesterol-rich microdomains and are thought to belong to signaling platforms involved in the induction of cell cycle arrest and cell death. Overall, this study reveals that mAb16D10 holds great potential to prevent pancreatic tumor proliferation by apoptotic cell death, thus promising therapeutic prospects for treatment of pancreatic adenocarcinoma, http://www.jimmunol.org/ a highly lethal disease. The Journal of Immunology, 2012, 189: 3386–3396.

n Western countries, pancreatic ductal adenocarcinoma (PDAC) or Ags. mAbs are clinically efficient: trastuzumab for advanced is the fourth leading cause of cancer death (1). PDAC resists chemotherapy of refractory breast cancer (6) and rituximab for I chemotherapy, radiotherapy, immunotherapy, and apoptotic treatment of non-Hodgkin’s lymphoma (7). mAbs have selective activators (2). Dysregulation of the apoptosis machinery and tumor-targeting potential associated with relatively low toxicity. evasion of apoptosis is a general mechanism in PDAC (3). Many mAbs deliver chemotherapeutic agents, target signaling mole- studies described mutations in oncogenes such as K-ras, a loss of cules, and induce complement-dependent cytotoxicity (CDC) or by guest on September 28, 2021 tumor suppressor such as p53, or dysregulated expression of Ab-dependent cellular cytotoxicity (ADCC) (8). mAbSC104 tar- proteins involved in the cell cycle, proliferation, and apoptosis gets colorectal cancer-associated glycolipids and induces tumor such as cyclins, glycogen synthase kinase-3b (GSK-3b), EGF-R, cell death in vitro independently of CDC or ADCC or block- or Bcl-2 (4, 5). Thus, adjuncts to treatment are needed. Thera- ade of known survival signaling pathways (9). mAbSC104 induces peutic strategies appear promising that involve tyrosine kinase apoptotic cell death and synergizes with chemotherapy in vivo inhibitors and mAbs targeting pancreatic tumor-specific receptors in mice. Ab-induced cell death may be crucial for treatment of solid tumors, which develop protection against CDC and ADCC (10). *Aix-Marseille Universite´, Centre de Recherche en Oncologie Biologique et Onco- In PDAC, a tumor-associated Ag referred to as feto-acinar pharmacologie, F-13005, Marseille, France; †INSERM, Unite´ Mixte de Recherche S 911, F-13005, Marseille, France; and ‡INSERM, Centre d’Investigation Clinique pancreatic protein (FAPP) was identified using mAbJ28 (11, 12). 9502, F-13005, Marseille, France We showed that FAPP is a glycoform of bile salt-dependent lipase Received for publication September 14, 2011. Accepted for publication July 23, (BSDL), whose glycosylation varies in pancreatic pathologies (13, 2012. 14). Onco-pathological BSDL-J28, as its physiological counter- This work was supported by grants from the Association pour la Recherche sur le part BSDL, is mainly localized at the plasma membrane level and Cancer (3816 to E.M.) and by institutional funding from INSERM and the Aix- Marseille Universite´. L.C. is a doctoral fellow of the Ministe`re de l’Enseignement within the endoplasmic reticulum and Golgi apparatus (15). We Supe´rieuretdelaRecherche,AssociationpourleDe´veloppement des Recherches then characterized another mouse mAb, mAb16D10, raised Biologiques et Me´dicales, and the Ligue contre le Cancer. against a tumor glycoform of BSDL isolated from pancreatic juice Address correspondence and reprint requests to Dr. Eric Mas, INSERM UMR 911– of a patient with pancreatic adenocarcinoma. mAbJ28 and Centre de Recherche en Oncologie Biologique et Oncopharmacologie, Campus Sante´-Timone, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 05, France. mAb16D10 recognize distinct glycotopes on the C-terminal do- E-mail address: [email protected] main of pathological glycoforms of BSDL (D. Lombardo and The online version of this article contains supplemental material. E. Mas, unpublished observations). mAb16D10 specifically rec- Abbreviations used in this article: ADCC, Ab-dependent cellular cytotoxicity; BSDL, ognizes tumor cell surface Ag of human pancreatic tissues, dis- bile salt-dependent lipase; CDC, complement-dependent cytotoxicity; CTB, chol- criminating human pancreatic tumor tissues from normal tissues era toxin subunit B; FAPP, feto-acinar pancreatic protein; GSK-3b, glycogen synthase kinase-3b;MbCD, methyl-b-cyclodextrin; PARP1, poly ADP-ribose polymerase including nontumor pancreas and from other cancer tissues (16). 1; PDAC, pancreatic ductal adenocarcinoma; PDMP, 1-phenyl-2-decanoyl- Moreover, mAb16D10 decreases the growth of pancreatic tumor amino-3-morpholino-1-propanol; PI, propidium iodide; siRNA, small interfering SOJ-6 cells xenografted in nude mice (17). This result suggests RNA. that this Ab could inhibit tumoral cell proliferation. Thus, we Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 sought first to test this possibility on different pancreatic tumor www.jimmunol.org/cgi/doi/10.4049/jimmunol.1102647 The Journal of Immunology 3387 cells in vitro and next to elucidate the intracellular pathways in- Expression of p53 mRNA induced by mAb16D10 volved in this process. We find that this novel Ab, which targets Total RNAwas isolated using TRIzol reagent (Invitrogen), and RT-PCR was a pancreatic glycoprotein, directly induces tumor cell apoptosis performed as described (18). p53 amplification was performed using a pair with a central involvement of GSK-3b signaling. of primers: forward, 59-ATGGAGGAGCCGCAGTCAGATCCTAGC-39; reverse, 59-TCAGTCTGAGTCAGGCCCTTCTGTCTTGA-39. Some specific PCR products were purified from agarose gel using the PCR Clean-up Materials and Methods Gel Extraction kit (Macherey-Nagel) and sequenced by Genome Express Reagents and Abs (Meylan, France). Patented mAb16D10 (IgM), which recognizes the O-glycosylated C- p53-targeted small interfering RNA treatment terminal domain of tumor BSDL, was homemade (17). Reagents were from these sources: Abs to Bax (sc-6236) and E-cadherin (H-108) (Santa SOJ-6 cells were transfected with p53-targeted small interfering RNA Cruz Biotechnology, Santa Cruz, CA); Abs to cleaved caspase-3 (Asp175), (siRNA) 1 and 2 (SignalSilence p53 siRNA Kit; Cell Signaling Technology) caspase-3, cleaved caspase-7 (Asp198), caspase-7, cleaved caspase-9 using Lipofectamine 2000 (Invitrogen). Transfection medium was replaced (Asp330), caspase-9, caspase-8, cleaved poly ADP-ribose polymerase 1 after 6 h with fresh 10% FCS medium before addition of mAb16D10 for 24 h. (PARP1) (Asp214), PARP1, Bcl-2, phospho–GSK-3b (Ser-9), total forms of GSK-3b and phospho-b-catenin (Ser33/37, Thr41) (Cell Signaling Cell cycle analysis Technology, Beverly, MA); Abs to caspase-8 (Abcam, Cambridge, U.K., b DNA contents were analyzed using PI as described (20). Cell cycle phase and Cell Signaling Technology); polyclonal Abs to -catenin (Abcam); distribution was detected by flow cytometry and analyzed by ModFit peroxidase-conjugated goat anti-rabbit IgG and anti-mouse IgG (respec- software (Verity Software House, Topsham, ME). The red fluorescence of tively Cell Signaling Technology and Calbiochem, San Diego, CA); Alexa- single event was recorded to measure the DNA index. conjugated goat anti-rabbit IgG, anti-mouse IgG and IgM (Invitrogen, Cergy-Pontoise, France); Abs to b-actin, irrelevant mouse IgM, FITC- SDS-PAGE and Western blotting Downloaded from conjugated goat anti-mouse IgM, fumonisin (B1 and B2), L-cycloserine, 1-phenyl-2-decanoylamino-3-morpholino-1-propanol (PDMP), methyl-b-cy- PAGE and Western blot analysis were performed as described (17). clodextrin (MbCD), filipin, MTT, propidium iodide (PI), and aphidicolin Membranes were developed with a chemiluminescent substrate (Roche (Sigma, St. Louis, MO); protease inhibitor mixture tablets (Roche Diag- Diagnostics). nostics, Meylan, France); RPMI 1640 and DMEM media, penicillin, streptomycin, trypsin–EDTA (Lonza, Le Vallois-Perret, France); and DAPI Statistical analysis

(Promega, Madison, WI). All data are mean 6 SD. Significant differences among the groups were http://www.jimmunol.org/ , Cell lines determined by unpaired Student t test. The p values 0.05 were con- sidered significant. Human pancreatic SOJ-6 tumor cells constitutively expressing 16D10 Ag were from Dr. M.J. Escribano (INSERM U260, Marseille, France). Human Results pancreatic tumor BxPC-3 and PANC-1 cells were from American Type Culture Collection (Rockville, MD). These cell lines were grown in 10% mAb16D10 decreases the number of living cells of only human heat-inactivated FCS medium (18). pancreatic tumor line-expressing surface 16D10 Ag Immunofluorescence SOJ-6 and BxPC-3 line cells expressed 16D10 Ag at their surface, as assessed by fluorocytometry analyses, whereas PANC-1 cells did not Cells were seeded in an 8-well culture plate (BD Falcon). Fresh 10% FCS by guest on September 28, 2021 medium was added 24 h before treatment with Abs. Cell membrane Ags (Fig. 1A). mAb16D10 reactivity decorated the plasma membranes were detected by indirect immunofluorescence (18) using a Carl Zeiss of the former cells, shown by fluorescence microscopy (Fig. 1B). microscope (Axiovert 200). Binding of mAb16D10 to its receptor led to internalization (data not shown). We next evaluated its ability to affect the viability of tumor Confocal microscopy cells in vitro. mAb16D10 significantly induced a decrease of mi- Cells were seeded on coverslips in 12-well plates (BD Falcon). All stages tochondrial dehydrogenase activity—an indicator of the number were performed at 4˚C. The cells were incubated successively with the of living cells—of SOJ-6 and BxPC-3 cells in a dose-dependent primary Abs for 1.5 h and then with secondary Ab coupled to Alexa 647 (Invitrogen) for 45 min. Ganglioside GM1 was detected with Vybrant manner (Fig. 2A). This decrease clearly became more important Alexa Fluor 488 Lipid Raft Labeling Kit (Invitrogen). The slides were at 48 h (data not shown). In parallel, relative PI-positive SOJ-6 and analyzed with a confocal microscope (Zeiss LSM 510). The images were BxPC-3 cells increased in a dose- and time-dependent manner (data processed with LAS AF Lite software. not shown). In further experiments, mAb16D10 was used at 25 mg/ Flow cytometry ml, which is close to the IC50. Moreover, we observed SOJ-6– and BxPC-3–decreased cell viability determined by blue trypan exclu- To quantify membrane Ags by flow cytometry, cells were harvested from sion after treatment with mAb16D10 (Fig. 2C, 2D). Notably, the culture plates with nonenzymatic cell-dissociation solution (Invitrogen) and labeled as described earlier. Cells were analyzed with a FACSCalibur device tumor cells, which have escaped death in the early phase, finally (BD Biosciences, le Pont-de-Claix, France). died by 72 h. In contrast, mAb16D10 had no effects on the viability of PANC-1 cells, whose plasma membranes were not reactive to this Cell viability assay Ab (Fig. 2A, 2E). The treatment of pancreatic tumor cells with ir- Mitochondrial succinate dehydrogenase activity, an indicator of the number relevant Abs did not significantly affect their growth (Fig. 2B). of viable cells, was determined by MTT assay (19). Cells were cultured in Altogether, these results indicate that the decrease of pancreatic 96-well culture plates for 24 h with 10% FCS and increasing concen- tumor cell viability was most likely mediated by the interaction trations of mAb16D10. Colorimetric signal was measured at 570 nm. Cell viability was also determined by trypan blue exclusion after treatment with between mAb16D10 and its cognate surface Ag. mAb16D10 at 0, 24, and 48 h. Independent experiments were performed at least three times; each point was replicated eight times. mAb16D10 evokes cell death and induces caspase activation Caspase activities We evaluated whether the decrease in cell viability and proliferation upon cell treatment with mAb16D10 was due to cell death. We Cells were treated with Abs in their respective medium with 10% FCS for therefore studied morphological changes of the nucleus. Chromatin 24 h before the addition of CaspACE FITC-VAD-FMK In Situ Marker according to the manufacturer’s instructions (Promega). Upon caspase action condensation was observed upon treatment of SOJ-6 and BxPC-3 cells on FITC-VAD-fmk, cells become fluorescent; these cells were counted on with mAb16D10 (Fig. 3A, arrows), contrary to mAb16D10-treated 10 fields randomly examined under fluorescence microscopy in triplicate. PANC-1 cells. The nucleus showed no morphological changes upon 3388 INDUCTION OF HUMAN PANCREATIC CANCER CELL DEATH BY mAb16D10 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 1. Expression of 16D10 Ag at the surface of human pancreatic tumor cells. (A) Cell surface 16D10 Ag expression on SOJ-6, BxPC-3, and PANC-1 cells was analyzed by flow cytometry using mAb16D10 (blue line) and irrelevant IgM (orange line) and compared with control (green line), which shows the nonspecific binding of secondary FITC-conjugated Abs. The mean fluorescence intensity of histograms is given in brackets. The percentages of the positive cell fractions for mAb16D10 (M1) were gated. (B) SOJ-6, BxPC-3, and PANC-1 cells were probed with mAb16D10 or irrelevant IgM before immunofluorescence microscopy analysis. Nonspecific binding of secondary FITC-conjugated Ab was used as internal control (original magnification 3200).

SOJ-6 and BxPC-3 cell treatment with irrelevant Abs (Fig. 3A). We observed (Fig. 3E) using two different specific Abs. An ultimate investigated whether the cell death promoted by mAb16D10 in- effector that is caspase-3– and caspase-7–dependent is PARP1. In- volved activation of cell death effector caspases. Caspase activation, activation of PARP1 resulted from its cleavage into fragments of 85 monitored by cell fluorescence, was significantly increased in SOJ-6 and 24 kDa. These fragments were detected in lysates of SOJ-6 cells and BxPC-3 cells after treatment with mAb16D10 compared with after treatment with mAb16D10 (Fig. 3D). These biochemical that in untreated cells and irrelevant Abs-treated cells (Fig. 3B). No changes associated with cell death support the notion that cell increase in fluorescent cells was detected in mAb16D10-treated treatment with mAb16D10 evoked cell death via the caspase-3–, PANC-1 cells compared with controls. Furthermore, mAb16D10 caspase-7–dependent pathway. induced a time-dependent increase in caspase enzymatic activities in SOJ-6 and BxPC-3 cells, with maximal numbers of fluorescent cells mAb16D10 induces p53 expression and affects Bax and Bcl-2 obtained at 24 h (Fig. 3C). To identify the caspases involved in this expression process, their cleavage was investigated upon treatment of SOJ-6 Because p53 tumor suppressor acts in proliferation, apoptosis, cellular cells with mAb16D10. Caspase-9, caspase-3, and caspase-7 were senescence, activation of DNA repair, and cell cycle arrest (21–24), we cleaved consecutively to mAb16D10 cell treatment, as assessed by investigated its expression in SOJ-6 cells with and without Ab immunoblotting (Fig. 3D). In contrast, no cleavage of caspase-8 was treatment. p53 was detected in SOJ-6 cells treated with mAb16D10 The Journal of Immunology 3389

FIGURE 2. mAb16D10 alters viability of human pancreatic tumor cells expressing 16D10 Ag. The cells were treated with mAb16D10 (A) or irrelevant IgM (B)at different concentrations for 24 h. The amount of living cells was evaluated by monitoring their mitochondrial respiratory chain activity using MTT assay. Results are expressed as percentage relative to controls Downloaded from (no Abs); data are means 6 SD of three independent experiments. *p , 0.05 (versus control). (C–E) Cell viability was evaluated using trypan blue exclusion. Abs were added at 0, 24, and 48 h. Results are expressed as number of living cells. Each http://www.jimmunol.org/ experiment was carried out in triplicate. by guest on September 28, 2021

but not in control SOJ-6 cells, untreated or treated with irrelevant expression was not modified in mAb16D10-treated PANC-1 cells Abs (Fig. 4A). This result was confirmed at the transcriptional level compared with controls. Hence, functional p53, Bax, and caspases 9, by RT-PCR using specific probes for p53 (Fig. 4B). Because the p53 3, and 7 play essential roles in the programmed death of tumor cells gene is mutated in up to 50% of pancreatic adenocarcinoma, RT- expressing 16D10 Ag. PCR p53 transcripts of SOJ-6 cells were sequenced. We found b a sequence coding for Arg72, which agrees with a report on the mAb16D10 promotes GSK-3 activation existence of two p53 isoforms with Arg or Pro at position 72 (25, 26) Cell treatment with mAb16D10 caused a G1/S arrest (G1/S peak: 96 and one mutation of Tyr236 to Asp236. To determine whether mu- versus 71% for control) (Fig. 5A, dashed area) and an increase of tated p53 yielded functional proteins in relation to the caspase- dead cells (sub-G1 peak: 6 versus 0.8%) (Fig. 5A, blue area), whereas dependent death pathway, we silenced its expression in SOJ-6 cells in G2/M phase decreased (G2/M peak: 4 versus 29%) (Fig. 5A, cells with two p53-targeted siRNAs. siRNA 2 was more efficient red area). These results were confirmed with cells synchronized in than siRNA 1 in inhibiting p53 expression (Fig. 4B). Next, to G1/S phase by aphidicolin. To note, mAb16D10 treatment had no evaluate mutated p53 functionality, we assessed caspase activity effect on the cell cycle of PANC-1 cells (data not shown). in SOJ-6 cells treated with mAb16D10 or not treated. Relative to Next, we examined the status of GSK-3b in mAb16D10-treated non-transfected cells, a significantly decreased percentage of fluo- SOJ-6 cells. This kinase affects expression of cyclin D1, a modu- rescent transfected cells, reflecting caspase cleavages triggered by lator of the cell cycle. Whereas cell treatment did not affect the mAb16D10 treatment, was noted with the two p53-targeted siRNAs expression of total GSK-3b, the Ser9-phosphorylated-GSK-3b (Fig. 4C). Again, siRNA 2 was more efficient than siRNA 1. We (inactive) form was decreased, contrary to controls (Fig. 5B). Fig. concluded that p53 was thus functional and required for caspase 5C shows that cell treatment decreased cyclin D1 expression, inactivation. versely to GSK-3b activation. The ability of p53 to promote cell death can be linked to its ability to A dramatic amount of phosphorylated b-catenin, which is phos- induce Bax transcription directly (27) and to alter that of Bcl-2 or its phorylated by GSK-3b, was detected in SOJ-6 cells treated with activity (28). Consistent with this, we found that both the expression mAb16D10, contrary to controls (Fig. 5C). To get insight into the of proapoptotic Bax protein and inhibition of expression of anti- GSK-3b functional implication, GSK-3b activity was inhibited with apoptotic Bcl-2 protein resulted from cell mAb16D10 treatment but lithium chloride. This inhibition reversed the effects triggered by not from irrelevant Ab or medium treatment (Fig. 4D). Bax/Bcl-2 mAb16D10 in SOJ-6 cells: we observed strongly decreased phos- 3390 INDUCTION OF HUMAN PANCREATIC CANCER CELL DEATH BY mAb16D10 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 3. mAb16D10 evokes cell death and induces caspase activation. (A) After treatment with Abs, cells were fixed and permeabilized with methanol for 5 min at –20˚C and stained with DAPI solution (2.5 mg/ml) for 1 min at room temperature. The nuclear morphology of cells was then observed using a fluorescent microscope with a 370-nm filter. Cells with chromatin condensation are indicated by arrows (original magnification 3400). (B) Various cell lines were treated or not with Abs for 24 h prior to the addition of 10 mM CaspACE FITC-VAD-FMK In Situ Marker in the culture medium. Results are expressed as mean 6 SD of fluorescent cell number/cm2 of three independent experiments. **p , 0.005 (versus control). (C) Kinetics of caspase activation in SOJ-6 and BxPC-3 cells. Results are expressed as percentage relative to the maximal number of fluorescent cells, taken as 100%. Data are fluorescence intensity mean 6 SD of three independent experiments. (D and E) SOJ-6 cells were treated with mAb16D10 or irrelevant IgM for 24 h. Cell lysate proteins (50 mgperlane)wereloadedonSDS-PAGE,transferredtoni- trocellulose, and probed with specific Abs as indicated. b-Actin was used as internal control. Representative immunoblot of three independent experiments. phorylation of b-catenin (Fig. 5D), inhibition of p53 expression (Fig. status. After treatment with mAb16D10, b-catenin was detected 5E), and a significantly decreased percentage of fluorescent cells mainly in the cytosolic compartment of SOJ-6 cells, under phos- with caspase cleavage (Fig. 5F). Notably, our data evidenced func- phorylated form, whereas total b-catenin was localized at the tional relationships between GSK-3b and both b-catenin and p53. plasma membrane of untreated cells (Fig. 6A). SOJ-6 cells constitutively do not exhibit detectable plasma b mAb16D10 alters -catenin localization and E-cadherin membrane E-cadherin. Upon mAb16D10 treatment, SOJ-6 cells expression markedly expressed E-cadherin at the plasma membrane (Fig. 6B). Next, we sought to determine whether treatment with mAb16D10 The ganglioside GM1, which is selectively partitioned in lipid could alter b-catenin localization together with phosphorylation microdomains, colocalized with E-cadherin in mAb16D10-treated The Journal of Immunology 3391 Downloaded from

A FIGURE 4. mAb16D10 induces expression of p53 and proapoptotic Bax and inhibits expression of antiapoptotic Bcl-2 in SOJ-6 cells. ( ) Tumor cells http://www.jimmunol.org/ were treated with mAb16D10 or irrelevant IgM for 24 h. Cell lysate proteins (50 mg per lane) were loaded on SDS-PAGE, transferred to nitrocellulose membrane, and probed with Abs to p53. b-Actin was used as internal control. Representative immunoblot of three independent experiments. (B) SOJ-6 cells were transfected with p53-targeted siRNA 1 or 2 and further incubated with or without mAb16D10 for 24 h. RT-PCR experiments were performed using specific primers for p53 tumor suppressor. b-Actin was taken as internal control. After PCR, aliquots of amplified products were electrophoresed on 1% agarose gel in the presence of ethidium bromide. The sizes of amplified products are indicated in base pairs. (C) Effects of mAb16D10 on p53-targeted siRNA-treated cells monitored by CaspACE FITC-VAD-FMK In Situ Marker. Results are expressed as percentage relative to the maximal number of fluorescent cells, taken as 100%. Data are fluorescence mean 6 SD of three independent experiments. *p , 0.06, **p , 0.01. (D) Tumor cells were treated with mAb16D10 or irrelevant IgM for 24 h. Bcl-2 and Bax expressions by SOJ-6 and PANC-1 tumor cells were probed with specific Abs. b-Actin was used as internal control. Representative immunoblot of three independent experiments. by guest on September 28, 2021

SOJ-6 cells (Fig. 6C). E-cadherin expression was confirmed by between mAb16D10 and 16D10 Ag on pancreatic tumor cells at immunoblots of cell lysates of SOJ-6 cells treated with the level of membrane lipid microdomains is illustrated in Fig. 8. mAb16D10 but not in lysates of irrelevant Ab- or mock-treated cells (Fig. 6D). PANC-1 cells, which did not exhibit membrane- Discussion 16D10 Ag (Fig. 1B), did not express E-cadherin at a detectable Resistance of pancreatic cancer to apoptotic cell death is a major level at their plasma membrane, with or without mAb16D10 factor preventing responses to current therapies. In this study, we treatment (Fig. 6B). Thus, E-cadherin expression at the plasma show that mAb16D10 is able to induce pancreatic tumoral cell cell membrane depended on both the presence of the 16D10 Ag at death. The decrease of cell viability induced by mAb16D10 the cell surface and its ligation to mAb16D10. requires 16D10 Ag expression at the tumor cell surface, as human PANC-1 cells, which do not express this Ag, are insensitive to Disorganization of membrane lipid microdomains decreases mAb16D10, contrary to other tumor cell lines that express 16D10 the antiproliferative effects of mAb16D10 Ag, such as SOJ-6 and BxPC-3. Moreover, cell cycle analysis Fig. 7A shows the 16D10 Ag colocalized with GM1 at the SOJ-6 indicates that tumor cells treated with mAb16D10 accumulate in cell surface. To address the role of lipid microdomains, we used the G1/S phase, like renal carcinoma cells treated with mAb to MbCD and filipin, drugs reported to deplete cholesterol in CD26 (20, 31). membrane rafts or to sequester cholesterol, respectively, without Altogether, our results reveal the pivotal involvement of GSK-3b altering cell viability (29). In the presence of MbCD and filipin, signaling in action mechanisms induced by mAb16D10, which the effects of mAb16D10 on the cell viability significantly de- rely on activation of cell death processes (p53 dependent) and on creased (Fig. 7B). Furthermore, MbCD decreased the expression downregulation of active b-catenin and cyclin D1. of the 16D10 Ag contrary to filipin (Table I; Supplemental Fig. 1). Dead and dying cells obtained after mAb16D10 treatment were Because sphingolipids also participate in the raft structure, assessed by chromatin condensation, significant activation of we used metabolic inhibitors of (glyco)sphingolipid biosynthesis caspases and marked proapoptotic Bcl-2 family protein (Bax) (15). Tested at efficient concentrations, neither L-cycloserine (an expression, and presence of sub-G1 DNA content. These mor- inhibitor of serine palmitoyltransferase) nor fumonisin 1 or 2 (in- phological, biochemical, and functional criteria serve to identify hibitors of dihydroceramide synthetase) interfered with the anti- cell death modalities as being apoptosis (32). In brief, caspase-9, proliferative effects of mAb16D10 (Fig. 7C). We next tested caspase-3, and caspase-7 activation, PARP1 cleavage, conspicuous PDMP, an inhibitor of glycosphingolipid synthesis, acting on the Bax expression, Bcl-2 inhibition, and p53 protein expression are last step of sphingolipid synthesis (30). PDMP partly reversed the found in mAb16D10-treated SOJ-6 cells; these results are con- effects of mAb16D10 on SOJ-6 cell proliferation. The interaction sistent with the facts that p53 can upregulate Bax (27), may di- 3392 INDUCTION OF HUMAN PANCREATIC CANCER CELL DEATH BY mAb16D10 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 5. mAb16D10 arrests cell cycle in G1/S phase and promotes GSK-3b activation in SOJ-6 cells. (A) Cell cycle distribution of SOJ-6 cells synchronized or not with aphidicolin (5 mg/ml) for 6 h, further incubated with or without mAb16D10 for 24 h. Each experiment was carried out in triplicate. (B and C) SOJ-6 cells were treated for 24 h with mAb16D10 or irrelevant IgM. Total GSK-3b, phospho–GSK-3b, cyclin D1, total b-catenin, and phospho–b-catenin expressions were detected in cell lysate by Western blot using speciﬁc Abs. b-Actin was used as internal control (50 mg total protein per lane). Representative immunoblot of three independent experiments. (D and E) SOJ-6 cells were treated with mAb16D10 or irrelevant IgM for 24 h after 1-h preincubation with 5 mM lithium chloride (LiCl), an inhibitor of GSK-3b activity. Representative immunoblot of p53 from whole-cell lysates showing the effects of LiCl (D)onb-catenin, (E) on p53 expression. b-Actin was used as internal control. (F) Functional effects of mAb16D10 on LiCl-treated cells monitored by CaspACE FITC-VAD-FMK In Situ Marker. Maximal number of ﬂuorescent cells was taken as 100%. Results are mean 6 SD of three independent experiments. ***p , 0.001. The Journal of Immunology 3393 Downloaded from

FIGURE 6. mAb16D10 alters b-catenin localization and phosphorylation and E-cadherin expression at the plasma membrane of tumor cells expressing 16D10 Ag. (A) SOJ-6 cells treated with mAb16D10 were labeled with Abs to total b-catenin and phospho–b-catenin before fluorescence microscope http://www.jimmunol.org/ analyses (original magnification 3400). (B) After treatment with mAb16D10, SOJ-6 and PANC-1 cells were labeled with Ab to E-cadherin (original magnification 3400). (C) SOJ-6 cells treated with mAb16D10 were probed with Ab to E-cadherin; ganglioside GM1 was stained with Alexa 488–cholera toxin subunit B (CTB) and cross-linked by anti-CTB according to the manufacturer’s instructions before confocal microscopy analyses (original magnification 3945). (D) SOJ-6 cells were treated with mAb16D10 or irrelevant IgM for 24 h, and cell lysate proteins were analyzed by Western blot using Abs to E-cadherin. b-Actin was used as internal control (50 mg total protein per lane). rectly activate the proapoptotic Bax in the cytosol to permeabilize vation of PARP1, which plays an essential role in repair of mitochondria membrane (33), and simultaneously downregulates damaged DNA (34), and finally trigger cell death (27, 33). The Bcl-2 (27). These events result in caspase activation and inacti- gene TP53 has a common sequence polymorphism in the proline- by guest on September 28, 2021

FIGURE 7. Ag 16D10 is localized in membrane lipid microdomains, and disorganization of these microdomains decreases the antiproliferative effects of mAb16D10. (A) SOJ-6 cells were probed with mAb16D10; ganglioside GM1 was stained with Alexa 488–CTB and cross-linked by anti-CTB before confocal microscopy analyses (original magniﬁcation 3945). (B and C) The culture medium of SOJ-6 cells was replaced by fresh medium containing MbCD or ﬁlipin (B) or metabolic inhibitors of glycosphingolipid biosynthesis at indicated concentration for 6 h (C). Next, the medium was replaced with fresh medium containing drugs plus mAb16D10 or irrelevant Abs for 24 h. The amount of living cells was determined by monitoring their mitochondrial respiratory chain activity using MTT assay. *p , 0.05, **p , 0.001 (versus control). Results are expressed as a percentage relative to controls without drugs; data are means 6 SD of three independent experiments. F1, Fumonisin 1; F2, fumonisin 2; LCS, L-cycloserine. 3394 INDUCTION OF HUMAN PANCREATIC CANCER CELL DEATH BY mAb16D10

Table I. Cell surface 16D10 Ag expression analyzed by flow cytometry nisms. Others reported that activated GSK-3b phosphorylates using mAb16D10 on SOJ-6 cells treated with MbCD and filipin b-catenin in the cytoplasm; once phosphorylated, it is degraded and unable to participate in regulation of transcription of target MbCD Variation Filipin Variation genes like cyclin D1 (28-40). Besides, activated GSK-3b trans- (mM) MFIa (%)b (mM) MFIa (%)b locates into the nucleus and phosphorylates cyclin D1, which is 0 1045 0 0 1045 0 afterward exported from the nucleus and degraded by the ubiq- 50 1099 5 0.5 1572 50 2 uitin-dependent proteasome complex (37, 38). Our data demon- 100 838 20 1 1511 45 b 500 641 239 5 1523 46 strate that after exposure to an inhibitor of GSK-3 activity, b a mAb16D10-treated SOJ-6 cells no longer express -catenin under Mean fluorescence intensity of cells stained with mAb16D10. b bPercentage of variation versus control; average of three independent experi- phosphorylated form. Consistent with the above studies, -catenin ments. therefore should be available to intervene in the regulation of transcription of cyclin D1 gene. GSK-3b has direct or indirect relationships with p53. Once rich domain of p53, which is necessary for the protein fully to activated, GSK-3b indirectly may negatively regulate p53 through induce apoptotic cell death (25). This results in either proline or phosphorylation and activation of Mdm2, an important negative arginine at amino acid position 72, the latter inducing apoptosis regulator of p53 (41). However, after DNA damage, GSK-3b may markedly better than the Pro72 variant. Notably, the p53 variant positively and directly interact with p53, in both nucleus and expressed by SOJ-6 cells contains Arg72. It also contains one mitochondria, contributing to the transcriptional and apoptotic mutation of Tyr236 to Asp236, which may affect the protein dy- actions of p53 (42). Therefore, we investigated the functional Downloaded from namic (35). Importantly, silencing of p53 expression in SOJ-6 interaction between transcriptional expression of p53 and GSK-3b cells significantly decreases cell death induced by mAb16D10, activation induced via the binding of the Ab to its cognate Ag. which substantiates the notion that mAb16D10 evokes cell death Two arguments suggest that p53-mediated caspase activation by activation of the p53-dependent apoptotic pathway. involves GSK-3b. The inhibition of GSK-3b activity by LiCl first Besides, activation of p53 leads to diverse cellular responses, markedly inhibited p53 expression suggesting that GSK-3b ac-

including cell cycle arrest involving effector proteins, in particular tivity and mAb16D10-induced p53 expression are linked. Second, http://www.jimmunol.org/ p21 for G1 arrest (23, 36), and thus might be involved in the cell it also led to dramatically decreased cell death. Together, these cycle arrest in the G1/S phase of mAb16D10-treated SOJ6 cells. results may imply that GSK-3b positively regulates p53 level and Also, treatment of tumor cells with mAb16D10 favors the acti- promotes its action. vation of GSK-3b, which substantially phosphorylates b-catenin, GSK-3b also regulates the interaction of b-catenin with E- associated with a markedly decreased cyclin D1 expression. cadherin (43). Serine residues in the E-cadherin cytoplasmic do- Cyclin D1, whose overexpression is most frequently associated main are phosphorylated by GSK-3b and casein kinase II, creating with human cancer (30), shortens the G1/S phase transition and additional interactions with b-catenin. In this study, we demon- promotes cell progression and differentiation (37). In this, de- strated that mAb16D10 treatment induces plasma membrane E- creased cyclin D1 expression may rely on nonexclusive mecha- cadherin expression in SOJ-6 cells, whereas intriguingly, b-cat- by guest on September 28, 2021

FIGURE 8. Schematic representation of mAb16D10-targeted signaling pathways toward cell death. mAb16D10 interacts with 16D10 Ag on pancreatic tumor cells at the level of membrane lipid microdomains. mAb16D10 requires a cholesterol-dependent intact structure of microdomains to trigger cell death. Via these raft signaling platforms, mAb16D10 leads to p53 and GSK-3b activations. GSK-3b positively regulates p53 level and promotes its action. Activated p53 induces Bax expression and downregulates Bcl-2, resulting in the apoptotic caspase pathway with inactivation of PARP. Both p53 and GSK- 3b downregulate cyclin D1, driving to cell cycle arrest. p53 tightly controls the expression of p21 gene. The p21 protein functions as a regulator of cell cycle progression at G1 and mediates the p53-dependent cell cycle G1 phase arrest in response to a variety of stress stimuli. GSK-3b may also indirectly act on cyclin D1 by inhibiting b-catenin. b-Catenin is phosphorylated and degraded. This process could happen after dissociation from its complex with E- cadherin, provoked by the binding of mAb with tumor 16D10 Ag, which promotes E-cadherin overexpression. Black line: signaling pathway known. Blue line: signaling pathway known and evidenced in this work. Dashed blue line: signaling pathway unknown and suggested by this work. The Journal of Immunology 3395 enin is mainly detected in the cytosolic compartment, in contrast of mAb16D10 to tumor Ag at the cell surface triggers destabili- to untreated SOJ-6 cells. Thus, b-catenin no longer seems to bind zation of b-catenin and E-cadherin complexes favoring the to E-cadherin and is translocated from the plasma membrane to phosphorylation of b-catenin by GSK-3b. This requires further the cytoplasm in mAb16D10-treated tumor cells. Moreover, the thorough investigations in such peculiar cells that are tumor cells. increased expression of phospho–b-catenin upon mAb16D10 cell To summarize, mAb16D10 exerts pleiotropic and converging treatment and its cytoplasmic translocation suggest further deg- effects on pancreatic cells expressing the 16D10 Ag to trigger cell radation via the proteasome (39, 44). death, illustrated in Fig. 8. The E-cadherin/b-catenin–mediated cell adhesion system is Inducing apoptosis of pancreatic cancer cells by targeting known to act as a suppressive system for invasion (45). When protein is a major goal for development of therapeutic agents. b-catenin is bound to E-cadherin, signaling-competent nuclear mAb16D10, which specifically recognizes tumor cell surface Ag b-catenin levels diminish, and cell proliferation and invasion are of human pancreatic tissues (16) and inhibits the development of suppressed (45). E-cadherin/b-catenin expression is lost in inva- xenografted human pancreatic tumor (17), causes death of human sive digestive cancers (46, 47); partial or complete loss of E- pancreatic tumor cells (this study). These data strongly suggest cadherin expression correlates with malignancy (48), whereas mAb16D10 may fight against PDAC. Moreover, the high expres- overexpression of E-cadherin leads to defective invasion of mel- sion of E-cadherin promoted by cell treatment with mAb16D10 anoma cells (49). Thus, alteration in cadherin/catenin function or would be beneficial to patients by leading to defective malignancy expression is found in the neoplastic process as a step in metas- of pancreatic tumor cells. Therefore, mAb16D10 represents a tasis (4). promising tool in the development of new Ab-based therapeutic

Our data are in line with the results of Leong’s et al. (50) strategies, using chimeric- or humanized-Ab versions, against Downloaded from showing that reexpression of E-cadherin in E-cadherin-negative pancreatic adenocarcinoma, hopefully reinforced by the immune breast tumors (due to inhibition of ligand-induced Notch signal- system through recognition of apoptotic tumor cells. ing) is associated with suppression of active b-catenin, with apoptosis and with both tumor growth and metastasis inhibition. Acknowledgments Notably, chronic exposure of A431 epidermoid carcinoma to We wish to thank Dr. P. Golstein (Centre d’Immunologie de Marseille- cetuximab induces an upregulation of E-cadherin expression (51). http://www.jimmunol.org/ Luminy, Marseille, France) for helpful comments. We are indebted to Because morbidity in most cancer patients is not due to primary C. Prevot (Flow Cytometry Core Facilities, INSERM UMR 911–Centre cancer but to metastatic disease, which is particularly true in de Recherche en Oncologie Biologique et Oncopharmacologie, Campus PDAC (1), the high expression of E-cadherin promoted by cell Sante´-Timone, Marseille, France) for help in analyses of flow cytometry treatment with mAb16D10 would be beneficial to patients by results, to Dr. L. Panicot-Dubois (INSERM UMR 911–Centre de leading to defective malignancy of pancreatic tumor cells. Recherche en Oncologie Biologique et Oncopharmacologie) for help in 16D10 Ag and E-cadherin, associated with b-catenin, are confocal microscopy, and to Dr. T. Pourcher for the generous gift of confined in lipid microdomains, in agreement with the finding that technical support (UMR E4320 Laboratoire TIRO, Nice, France). BSDL is associated with intracellular rafts during transportation to cell membrane (15). Notably, mAb16D10 binding seems to re- Disclosures by guest on September 28, 2021 quire a cholesterol-dependent intact structure of lipid micro- The authors have no financial conflicts of interest. domains to trigger cell death as both depletion in cholesterol and/ or Ag and sequestration of cholesterol associated with a marked Ag expression lead to reversion of antiproliferative effects of References mAb1610. The association of 16D10 Ag with membrane lipid 1. Jemal, A., R. Siegel, E. Ward, Y. Hao, J. Xu, T. Murray, and M. J. Thun. 2008. microdomains occurs once these domains are structured because Cancer statistics, 2008. CA Cancer J. Clin. 58: 71–96. early inhibitors of sphingolipid synthesis are inefficient in re- 2. Li, X., A. B. Roginsky, X. Z. Ding, C. Woodward, P. Collin, R. A. Newman, R. H. Bell, Jr., and T. E. Adrian. 2008. Review of the apoptosis pathways in versing mAb16D10 effects. This inefficiency also suggests that the pancreatic cancer and the anti-apoptotic effects of the novel sea cucumber salvage pathway recycling surface glycosphingolipids (52) is compound, Frondoside A. Ann. N. Y. Acad. Sci. 1138: 181–198. 3. Hamacher, R., R. M. Schmid, D. Saur, and G. Schneider. 2008. Apoptotic sufficient to ensure the formation of lipid rafts to which the 16D10 pathways in pancreatic ductal adenocarcinoma. Mol. Cancer 7: 64. Ag associates. Only PDMP, which blocks ganglioside synthesis 4. Garcea, G., M. M. Manson, C. P. Neal, C. J. Pattenden, C. D. Sutton, (53), partially decreased cell death induced by mAb16D10. This A. R. Dennison, and D. P. Berry. 2007. Glycogen synthase kinase-3 beta; a new target in pancreatic cancer? Curr. Cancer Drug Targets 7: 209–215. might be because the inhibitor impairs the intracellular pathway of 5. Giovannetti, E., V. Mey, S. Nannizzi, G. Pasqualetti, M. Del Tacca, and BSDL (15) and/or that the drug affects the association of the R. Danesi. 2006. Pharmacogenetics of anticancer drug sensitivity in pancreatic 16D10 Ag-bearing protein with lipid rafts, as for the ErbB2 Ag cancer. Mol. Cancer Ther. 5: 1387–1395. 6. Yip, Y. L., and R. L. Ward. 2002. Anti-ErbB-2 monoclonal antibodies and ErbB- (54). Thus, the integrity of cholesterol-rich lipid microdomains 2-directed vaccines. Cancer Immunol. Immunother. 50: 569–587. might be a prerequisite to the presence of the 16D10 Ag at the 7. Grillo-Lo´pez, A. J., E. Hedrick, M. Rashford, and M. Benyunes. 2002. Rituximab: ongoing and future clinical development. Semin. Oncol. 29(1, Suppl 2)105–112. surface of pancreatic tumor cells. 8. Laheru, D., and E. M. Jaffee. 2005. Immunotherapy for pancreatic cancer - Roitbak et al. (55) showed that b-catenin and E-cadherin science driving clinical progress. Nat. Rev. Cancer 5: 459–467. complexes are associated with the polycystins and the raft 9. Durrant, L. G., S. J. Harding, N. H. Green, L. D. Buckberry, and T. Parsons. 2006. A new anticancer glycolipid monoclonal antibody, SC104, which directly marker flotillin-2 in primary human kidney epithelial cells. These induces tumor cell apoptosis. Cancer Res. 66: 5901–5909. molecules might confer to rafts the role of signaling platforms. 10. Gorter, A., and S. Meri. 1999. Immune evasion of tumor cells using membrane- This notion is supported by substantial evidence that raft integrity bound complement regulatory proteins. Immunol. Today 20: 576–582. 11. Escribano, M. J., J. Cordier, M. Nap, F. J. Ten Kate, and P. Burtin. 1986. Dif- is crucial for the initiation and maintenance of intracellular signals ferentiation antigens in fetal human pancreas. Reexpression in cancer. Int. J. (56, 57). Thus, E-cadherin, associated with b-catenin, and 16D10 Cancer 38: 155–160. 12. Escribano, M. J., and S. Imperial. 1989. Purification and molecular character- Ag in lipid rafts might therefore belong to signaling platforms ization of FAP, a feto-acinar protein associated with the differentiation of human involved in the induction of cell cycle arrest and cell death. At pancreas. J. Biol. Chem. 264: 21865–21871. present, we do not know the exact relationships between tumor Ag 13. Mas, E., N. Abouakil, S. Roudani, F. Miralles, O. Guy-Crotte, C. Figarella, b M. J. Escribano, and D. Lombardo. 1993. Human fetoacinar pancreatic protein: and the complex E-cadherin– -catenin and signal transductions an oncofetal glycoform of the normally secreted pancreatic bile-salt-dependent involving p53 and GSK-3b. The prime mover may be that binding lipase. Biochem. J. 289: 609–615. 3396 INDUCTION OF HUMAN PANCREATIC CANCER CELL DEATH BY mAb16D10

14. Lombardo, D. 2001. Bile salt-dependent lipase: its pathophysiological implica- 35. Can˜adillas, J. M., H. Tidow, S. M. Freund, T. J. Rutherford, H. C. Ang, and tions. Biochim. Biophys. Acta 1533: 1–28. A. R. Fersht. 2006. Solution structure of p53 core domain: structural basis for its 15. Aubert-Jousset, E., N. Garmy, V. Sbarra, J. Fantini, M. O. Sadoulet, and instability. Proc. Natl. Acad. Sci. USA 103: 2109–2114. D. Lombardo. 2004. The combinatorial extension method reveals a sphingolipid 36. el-Deiry, W. S., T. Tokino, V. E. Velculescu, D. B. Levy, R. Parsons, J. M. Trent, binding domain on pancreatic bile salt-dependent lipase: role in secretion. D. Lin, W. E. Mercer, K. W. Kinzler, and B. Vogelstein. 1993. WAF1, a potential Structure 12: 1437–1447. mediator of p53 tumor suppression. Cell 75: 817–825. 16. Benkoe¨l, L., J. P. Bernard, M. J. Payan-Defais, L. Crescence, C. Franceschi, 37. Diehl, J. A., M. Cheng, M. F. Roussel, and C. J. Sherr. 1998. Glycogen synthase M. Delmas, M. Ouaissi, B. Sastre, J. Sahel, A. M. Benoliel, et al. 2009. kinase-3beta regulates cyclin D1 proteolysis and subcellular localization. Genes Monoclonal antibody 16D10 to the COOH-terminal domain of the feto-acinar Dev. 12: 3499–3511. pancreatic protein targets pancreatic neoplastic tissues. Mol. Cancer Ther. 8: 38. Takahashi-Yanaga, F., and T. Sasaguri. 2008. GSK-3beta regulates cyclin D1 282–291. expression: a new target for chemotherapy. Cell. Signal. 20: 581–589. 17. Panicot-Dubois, L., M. Aubert, C. Franceschi, E. Mas, F. Silvy, C. Crotte, 39. Aberle, H., A. Bauer, J. Stappert, A. Kispert, and R. Kemler. 1997. beta-catenin J. P. Bernard, D. Lombardo, and M. O. Sadoulet. 2004. Monoclonal antibody is a target for the ubiquitin-proteasome pathway. EMBO J. 16: 3797–3804. 16D10 to the C-terminal domain of the feto-acinar pancreatic protein binds to 40. Liu, C., Y. Li, M. Semenov, C. Han, G. H. Baeg, Y. Tan, Z. Zhang, X. Lin, and membrane of human pancreatic tumoral SOJ-6 cells and inhibits the growth of X. He. 2002. Control of beta-catenin phosphorylation/degradation by a dual- tumor xenografts. Neoplasia 6: 713–724. kinase mechanism. Cell 108: 837–847. 18. Mas, E., E. Pasqualini, N. Caillol, A. El Battari, C. Crotte, D. Lombardo, and 41. Kulikov, R., K. A. Boehme, and C. Blattner. 2005. Glycogen synthase kinase 3- M. O. Sadoulet. 1998. Fucosyltransferase activities in human pancreatic tissue: dependent phosphorylation of Mdm2 regulates p53 abundance. Mol. Cell. Biol. comparative study between cancer tissues and established tumoral cell lines. 25: 7170–7180. Glycobiology 8: 605–613. 42. Watcharasit, P., G. N. Bijur, L. Song, J. Zhu, X. Chen, and R. S. Jope. 2003. 19. Mosmann, T. 1983. Rapid colorimetric assay for cellular growth and survival: ap- Glycogen synthase kinase-3beta (GSK3beta) binds to and promotes the actions plication to proliferation and cytotoxicity assays. J. Immunol. Methods 65: 55–63. of p53. J. Biol. Chem. 278: 48872–48879. 20. Ohnuma, K., T. Ishii, S. Iwata, O. Hosono, H. Kawasaki, M. Uchiyama, 43. Lickert, H., A. Bauer, R. Kemler, and J. Stappert. 2000. Casein kinase II H. Tanaka, T. Yamochi, N. H. Dang, and C. Morimoto. 2002. G1/S cell cycle phosphorylation of E-cadherin increases E-cadherin/beta-catenin interaction and arrest provoked in human T cells by antibody to CD26. Immunology 107: 325– strengthens cell-cell adhesion. J. Biol. Chem. 275: 5090–5095. 333. 44. Orford, K., C. Crockett, J. P. Jensen, A. M. Weissman, and S. W. Byers. 1997. Downloaded from 21. Vogelstein, B., D. Lane, and A. J. Levine. 2000. Surﬁng the p53 network. Nature Serine phosphorylation-regulated ubiquitination and degradation of beta-catenin. 408: 307–310. J. Biol. Chem. 272: 24735–24738. 22. Levine, A. J., W. Hu, and Z. Feng. 2006. The P53 pathway: what questions 45. Wong, A. S., and B. M. Gumbiner. 2003. Adhesion-independent mechanism for remain to be explored? Cell Death Differ. 13: 1027–1036. suppression of tumor cell invasion by E-cadherin. J. Cell Biol. 161: 1191–1203. 23. Hermeking, H. 2007. p53 enters the microRNA world. Cancer Cell 12: 414–418. 46. Chen, H. C., R. Y. Chu, P. N. Hsu, P. I. Hsu, J. Y. Lu, K. H. Lai, H. H. Tseng, 24. Brady, C. A., D. Jiang, S. S. Mello, T. M. Johnson, L. A. Jarvis, M. M. Kozak, N. H. Chou, M. S. Huang, C. J. Tseng, and M. Hsiao. 2003. Loss of E-cadherin D. Kenzelmann Broz, S. Basak, E. J. Park, M. E. McLaughlin, et al. 2011. expression correlates with poor differentiation and invasion into adjacent organs

Distinct p53 transcriptional programs dictate acute DNA-damage responses and in gastric adenocarcinomas. Cancer Lett. 201: 97–106. http://www.jimmunol.org/ tumor suppression. Cell 145: 571–583. 47. Li, Y. J., and X. R. Ji. 2003. Relationship between expression of E-cadherin- 25. Dumont, P., J. I. Leu, A. C. Della Pietra, III, D. L. George, and M. Murphy. 2003. catenin complex and clinicopathologic characteristics of pancreatic cancer. The codon 72 polymorphic variants of p53 have markedly different apoptotic World J. Gastroenterol. 9: 368–372. potential. Nat. Genet. 33: 357–365. 48. Mărgineanu, E., C. E. Cotrutz, and C. Cotrutz. 2008. Correlation between E- 26. Jeong, B. S., W. Hu, V. Belyi, R. Rabadan, and A. J. Levine. 2010. Differential cadherin abnormal expressions in different types of cancer and the process of levels of transcription of p53-regulated genes by the arginine/proline polymor- metastasis. Rev. Med. Chir. Soc. Med. Nat. Iasi 112: 432–436. phism: p53 with arginine at codon 72 favors apoptosis. FASEB J. 24: 1347–1353. 49. Molina-Ortiz, I., R. A. Bartolome´,P.Herna´ndez-Varas, G. P. Colo, and 27. Miyashita, T., and J. C. Reed. 1995. Tumor suppressor p53 is a direct tran- J. Teixido´. 2009. Overexpression of E-cadherin on melanoma cells inhibits scriptional activator of the human bax gene. Cell 80: 293–299. chemokine-promoted invasion involving p190RhoGAP/p120ctn-dependent in- 28. Hemann, M. T., and S. W. Lowe. 2006. The p53-Bcl-2 connection. Cell Death activation of RhoA. J. Biol. Chem. 284: 15147–15157. Differ. 13: 1256–1259. 50. Leong, K. G., K. Niessen, I. Kulic, A. Raouf, C. Eaves, I. Pollet, and A. Karsan.

29. Chen, X., and M. D. Resh. 2002. Cholesterol depletion from the plasma mem- 2007. Jagged1-mediated Notch activation induces epithelial-to-mesenchymal by guest on September 28, 2021 brane triggers ligand-independent activation of the epidermal growth factor re- transition through Slug-induced repression of E-cadherin. J. Exp. Med. 204: ceptor. J. Biol. Chem. 277: 49631–49637. 2935–2948. 30. Diehl, J. A., F. Zindy, and C. J. Sherr. 1997. Inhibition of cyclin D1 phosphor- 51. Oliveras-Ferraros, C., A. Vazquez-Martin, S. Cufı´, B. Queralt, L. Baéz, ylation on threonine-286 prevents its rapid degradation via the ubiquitin- R. Guardenõ, X. Hernańdez-Yaguë, B. Martin-Castillo, J. Brunet, and proteasome pathway. Genes Dev. 11: 957–972. J. A. Menendez. 2011. Stem cell property epithelial-to-mesenchymal transition 31. Inamoto, T., T. Yamochi, K. Ohnuma, S. Iwata, S. Kina, S. Inamoto, is a core transcriptional network for predicting cetuximab (Erbitux ™) efficacy in M. Tachibana, Y. Katsuoka, N. H. Dang, and C. Morimoto. 2006. Anti-CD26 KRAS wild-type tumor cells. J. Cell. Biochem. 112: 10–29. monoclonal antibody-mediated G1-S arrest of human renal clear cell carcinoma 52. Tettamanti, G., R. Bassi, P. Viani, and L. Riboni. 2003. Salvage pathways in Caki-2 is associated with retinoblastoma substrate dephosphorylation, cyclin- glycosphingolipid metabolism. Biochimie 85: 423–437. dependent kinase 2 reduction, p27(kip1) enhancement, and disruption of bind- 53. Lefrancois, S., T. May, C. Knight, D. Bourbeau, and C. R. Morales. 2002. The ing to the extracellular matrix. Clin. Cancer Res. 12: 3470–3477. lysosomal transport of prosaposin requires the conditional interaction of its 32. Kroemer, G., L. Galluzzi, P. Vandenabeele, J. Abrams, E. S. Alnemri, highly conserved d domain with sphingomyelin. J. Biol. Chem. 277: 17188– E. H. Baehrecke, M. V. Blagosklonny, W. S. El-Deiry, P. Golstein, D. R. Green, 17199. et al; Nomenclature Committee on Cell Death 2009. 2009. Classification of cell 54. Sottocornola, E., R. Misasi, V. Mattei, L. Ciarlo, R. Gradini, T. Garofalo, death: recommendations of the Nomenclature Committee on Cell Death 2009. B. Berra, I. Colombo, and M. Sorice. 2006. Role of gangliosides in the asso- Cell Death Differ. 16: 3–11. ciation of ErbB2 with lipid rafts in mammary epithelial HC11 cells. FEBS J. 33. Chipuk, J. E., T. Kuwana, L. Bouchier-Hayes, N. M. Droin, D. D. Newmeyer, 273: 1821–1830. M. Schuler, and D. R. Green. 2004. Direct activation of Bax by p53 mediates 55. Roitbak, T., Z. Surviladze, R. Tikkanen, and A. Wandinger-Ness. 2005. A poly- mitochondrial membrane permeabilization and apoptosis. Science 303: 1010– cystin multiprotein complex constitutes a cholesterol-containing signalling micro- 1014. domain in human kidney epithelia. Biochem. J. 392: 29–38. 34. Bryant, H. E., N. Schultz, H. D. Thomas, K. M. Parker, D. Flower, E. Lopez, 56. Deans, J. P., H. Li, and M. J. Polyak. 2002. CD20-mediated apoptosis: signalling S. Kyle, M. Meuth, N. J. Curtin, and T. Helleday. 2005. Specific killing of through lipid rafts. Immunology 107: 176–182. BRCA2-deficient tumours with inhibitors of poly(ADP-ribose) polymerase. 57. Patra, S. K. 2008. Dissecting lipid raft facilitated cell signaling pathways in Nature 434: 913–917. cancer. Biochim. Biophys. Acta 1785: 182–206.