Cancer Research Meeting Calendar Metabolism
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pelletier, L.
Right arrow Articles by Manié, S. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pelletier, L.
Right arrow Articles by Manié, S. N.
[Cancer Research 66, 3681-3687, April 1, 2006]
© 2006 American Association for Cancer Research

Cell, Tumor, and Stem Cell Biology

{gamma}-Secretase-Dependent Proteolysis of CD44 Promotes Neoplastic Transformation of Rat Fibroblastic Cells

Ludivine Pelletier1, Patricia Guillaumot1, Barbara Frêche1, Céline Luquain1, Dale Christiansen2, Sabine Brugière4, Jérome Garin3 and Serge N. Manié1

1 Génétique moléculaire, Signalisation et Cancer, UMR 5201, Faculté de Médecine, Lyon Cedex 08, France; 2 The Austin Research Institute, Austin Health, Heidelberg, Victoria, Australia; 3 Laboratoire de Chimie des Protéines, ERM 201 INSERM/CEA/UJF, Grenoble Cedex 9, France; and 4 Société Génome Express, Meylan, France

Requests for reprints: Serge N. Manié, Génétique moléculaire, Signalisation et Cancer, UMR 5201, Faculte de Médecine, 8 Avenue Rockefeller, 69 373 Lyon Cedex 08, France. Phone: 33-4-78-77-72-72; Fax: 33-4-78-77-72-20; E-mail: manie{at}sante.univ-lyon1.fr.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
The metalloprotease-dependent extracellular domain cleavage of the adhesion molecule CD44 is frequently observed in human tumors and is thought to promote metastasis. This cleavage is followed by {gamma}-secretase-dependent release of CD44 intracellular domain (CD44-ICD), which exhibits nuclear signaling activity. Using a reversible Ret-dependent oncogenic conversion model and a restricted proteomic approach, we identified a positive correlation between the neoplastic transformation of Rat-1 cells and the expression of standard CD44. In these transformed cells, CD44 was found to undergo a sequential metalloprotease and {gamma}-secretase cleavage, resulting in an increase in expression of CD44-ICD. We showed that this proteolytic fragment possesses a transforming activity. In support of this role, a significant and specific reduction in Ret-induced transformation of Rat-1 cells was observed following drug-mediated inhibition of {gamma}-secretase. Taken together, these findings suggest that the shedding of CD44 may not only modulate metastasis but also affects earlier events in tumorigenesis through the release of CD44-ICD. (Cancer Res 2006; 66(7): 3681-7)


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
CD44 is a widely distributed cell surface adhesion molecule that is implicated in a variety of physiologic and pathologic processes, including lymphocyte activation, cell-matrix interactions, and regulation of tumor growth and metastasis (see ref. 1 for a review). The gene that encodes CD44 consists of 20 exons. Alternative splicing of 10 so-called variable exons and different posttranslational modifications generates a large number of CD44 isoforms. The smallest and most widely expressed form is standard CD44, which lacks the entire variable region. Although the extracellular matrix component hyaluronic acid (HA) is the main ligand of CD44, several other molecules can also interact with this protein (2). In response to ligand binding, CD44 induces the transduction of various intracellular signals via both direct and indirect pathways (1). The organization of the signaling cascade is in part mediated by the association of CD44 intracellular domain (CD44-ICD) with the actin cytoskeleton via binding to ezrin, radixin, and moesin proteins.

The mechanisms by which CD44 plays a role in tumor progression are not yet clear although they do seem to be to implicated in cell survival signaling as well as regulation of cellular invasion and metastasis (1, 3). The proteolytic cleavage of membrane proteins, including CD44, has recently emerged as a key mechanism underlying their functional regulation. The shedding of CD44 after metalloprotease-dependent cleavage of the extracellular domain has been proposed as a mechanism to regulate cell detachment from HA (4, 5). Interestingly, the remaining membrane-anchored COOH-terminal fragment, termed CD44-CTF, has been frequently detected in brain, breast, lung, colon, and ovarian tumors (6). This suggests that the cleavage event may contribute to the oncogenesis of human tumors. Besides the shedding of CD44, the protein also undergoes regulated intramembrane proteolysis resulting in the release of CD44-ICD. Generation of CD44-ICD requires a presenilin-dependent {gamma}-secretase activity (7, 8) and leads to nuclear translocation of the liberated fragment, where it can cooperate with p300/cAMP-responsive element binding protein–binding protein to activate transcription (9). These findings suggest that CD44-ICD may function as an accessory modulator of transcriptional regulation, although to date the specifically regulated genes have not been identified. This scenario is remarkably similar to the cleavage of some other proteins among which the protein Notch is the most studied (10, 11). The intracellular cleavage of CD44 represents a novel aspect of signal transduction via CD44.

In a previous study, we described a controlled homodimerization of the receptor tyrosine kinase Ret, which mimics oncogenic activation responsible for the development of multiple endocrine neoplasia type 2A (MEN2A; ref. 12). Here, we took advantage of this reversible transformation system of immortalized rodent fibroblastic cells and found in a restricted proteomic analysis that standard CD44 is one of the main up-regulated protein associated with detergent-resistant membranes (DRM) during the transformation process. Moreover, CD44 undergoes a metalloprotease and subsequent regulated intramembrane proteolysis–dependent cleavage in Ret-transformed fibroblastic cells. Importantly, CD44-ICD was found to actively participate in the process of Ret-dependent transformation. Therefore, this study suggests a novel role for CD44 proteolytic cleavage in the oncogenesis of CD44-expressing tumors.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Antibodies and reagents. Peroxidase-conjugated antimouse or antirabbit antibodies were from Jackson ImmunoResearch (West Grove, PA). Anti-flotillin-2 was purchased from BD Transduction Laboratories (Franklin Lakes, NJ). Anti–green fluorescent protein (GFP) was from Roche Diagnostics (Mannheim, Germany). Streptavidin-horseradish peroxidase (HRP) was from Amersham (Piscataway, NJ). {gamma}-Secretase inhibitors (DAPT and L-685,458) and bovine testis hyaluronidase were from Calbiochem (La Jolla, CA). BB94 was from British Biotech (Oxford, United Kingdom). Anti-Erk and anti-phospho-Ret were from Cell Signaling Technology (Beverly, MA). Anti-CD44 and anti-Ret were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA).

DNA constructs, directed mutagenesis, and sequence analysis. Gateway Technology (Invitrogen, Carlsbad, CA) was used for cloning full-length CD44 protein (CD44-FL) and CD44-ICD. CD44-FL and CD44-ICD were amplified by PCR from cDNA of Rat-1 cell line using the following primers: forward, CD44-FL, 5'-ggggacaagtttgtacaaaaaagcaggcttcaccATGGACAAGGTTTGGTGGC-3'; ICD, 5'-ggggacaagtttgtacaaaaaagcaggcttcaccATGGCTGTCAACAGTAGGAGAAGGT-3'; reverse, 5'-ggggaccactttgtacaagaaagctgggtyCACCCCAATCTTCATATCCAC-3'. Small letters of the primers denote Att site. The PCR products were subcloned into the expression vector pDEST47 (Invitrogen) according to recommendations of the manufacturer. mRNAs were extracted from wild-type (WT) cells and MEN2A clones 1 and 2, using the Nucleospin RNA II (Macherey-Nagel, Duren, Germany) and subjected to Expend reverse transcriptase (Invitrogen). cDNAs were then used as template in PCR and the oligonucleotide primers synthesized for CD44-FL. PCR products were sequenced and analyzed. CD44 cDNA cysteine mutant was constructed by site-directed mutagenesis (QuickChange, Stratagene, La Jolla, CA) and subcloned into pDEST47. To substitute cysteine C289 and C298 with alanine, PCR-based site-directed mutagenesis was done using a set of primers including point mutations (underlined) as follows: 5'-TCTTGCCGTCGCCATTGCTGTCAACAGTAGGAGAAGGGCCGGGCAGAAGAAGAAGC-3' and 5'-GCTTCTTCTTCTGCCCGGCCCTTCTCCTACTGTTGACAGCAATGGCGACGGCAAGA-3'. PCR product was confirmed by DNA sequencing.

Cell culture, immunoblotting, and soft agar assay. Rat-1 clones that stably express Ret-WT, Ret-MEN2A, or Ret-Fv were maintained in complete medium (DMEM with 10% fetal bovine serum and antibiotics, all from Sigma, St. Louis, MO) supplemented with puromycin (2 µg/mL). Cells were further transfected with expression vectors containing CD44, cysteine-mutated CD44-FL (CD44-FL-C), or CD44-ICD cDNAs using Exgen 500 (Euromedex, Souffelweyersheim, France) according to recommendations of the manufacturer and selected in complete medium containing puromycin (2 µg/mL) and G418 (1 mg/mL). For drug treatment, cells were incubated overnight with 5 µmol/L BB94, 1 µmol/L DAPT, or 10 µmol/L L-685,458. Thereafter, cells were then treated with or without 100 units/mL hyaluronidase for 4 hours at 37°C. Cells were scrapped, washed with 1x PBS, and lysed at 4°C for 25 minutes in modified 1x radioimmunoprecipitation assay buffer [50 mmol/L Tris-HCl (pH 7.4), 1% NP40, 0.25% sodium deoxycholate, 150 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L NaOVO4, 1 mmol/L NaF, and protease inhibitors]. Precleared lysates (10 minutes at 13,000 rpm at 4°C) were subjected to SDS-PAGE under reducing conditions and Western blotting using the relevant probe, as previously described (13). Densitometry of the blots was assessed using a Fluo-S multi-imager system (Bio-Rad, Hemel Hempstead, United Kingdom). For anchorage-independent growth assay, 103 cells were seeded into a 0.35% agar medium containing complete medium onto a basal layer of 0.7% agar medium. When required, cells were treated with AP20187 (AP) or {gamma}-secretase inhibitors every 3 days as indicated. Cells were stained after 3 weeks and macroscopic colonies (>20 cells) from triplicate of each condition were scored.

DRM assay and purification of biotinylated proteins. Cell surface proteins were biotinylated with Sulfo-NHS-Biotin (Pierce Biotechnology, Rockford, IL) according to instructions of the manufacturer. Samples containing biotinylated proteins were submitted to a DRM assay (see below) and the fractions containing DRMs were pooled and diluted in TNE buffer. The mixture was centrifuged for 1 hour at 39,000 rpm in a SW41 rotor. To dissolve DRMs, pellets were suspended in TNE buffer containing 1% SDS and warmed up for 5 minutes at 65°C. Then, samples were diluted with TNE buffer containing 1% Triton X-100. Biotinylated proteins were purified by precipitation with immobilized avidin (Pierce Biotechnology). The beads were washed five times with TNE buffer containing 1% Triton and boiled for 5 minutes in the presence of 2% SDS and reducing agents. Released proteins were precipitated using a mixture of 4 volumes of methanol, 1 volume of chloroform, and 3 volumes of water and centrifuged for 5 minutes at 13,000 rpm. The lower layers were pooled and 3 volumes of methanol were added. Samples were centrifuged for 10 minutes at 13,000 rpm. Supernatant were discarded and the pellets were dried up at 30°C. Protein pellets were resuspended in Laemmli buffer containing reducing agents. Samples were resolved on a 10% SDS-PAGE gel of 0.75 mm, fixed with 30% ethanol-7.5% acetic acid, and stained with Coomassie blue for 1 hour. The gel was then destained and the protein bands were excised and analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). For DRM assay, cells were scrapped and lysed at 3 mg proteins/mL in ice-cold TNE buffer [150 mmol/L NaCl, 25 mmol/L Tris-HCl (pH 7.2), 4 mmol/L NaVO4, 5 mmol/L EGTA, and protease inhibitors] containing 0.5% Triton X-100 (Sigma) for 25 minutes at 4°C. Then, DRMs were isolated using a flotation assay based on buoyancy at low-density fractions of a bottom-loaded discontinuous sucrose gradient (12).

LC-MS/MS determination of protein sequences. Discrete bands were excised from the Coomassie blue–stained gel and subjected to tryptic digestion as described (14). Samples were injected into a nano-LC system directly coupled to a QTOF Ultima mass spectrometer (Waters, Milford, MA). MS and MS/MS data were acquired and processed automatically using Masslynx 4.0 software. Database searching was carried out using Mascot 2.0 (www.matrixscience.com). All the peptide sequences were checked manually.

GFP fluorescence microscopy. For analysis of the subcellular localization of CD44-ICD, cells were fixed with 4% paraformaldehyde for 10 minutes at 4°C and with methanol for 5 minutes at –20°C. Confocal microscopy was done with a Zeiss Axioplan 2 and 63x objective lens.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Rat-1 cell transformation induced by controlled homodimerization of Ret can be reverted. We have previously reported a controlled homodimerization of Ret that mimics oncogenic MEN2A activation (12). Briefly, this system relies on a chimeric Ret-Fv protein that can be activated in the presence of a synthetic bivalent dimerizing ligand AP. Immortalized Rat-1 fibroblastic clones either stably expressing Ret-Fv and chronically stimulated with AP, or stably expressing Ret carrying a representative MEN2A mutation at Cys634, show comparable anchorage-independent growth capability (Fig. 1, top ). This indicates that both Ret activation systems are equally potent in inducing Rat-1 cell transformation. AP-activated Ret offers the advantage of being reversible, whereas Ret MEN2A mutations are not. Thus, we took advantage of this controlled homodimerization to address whether the cellular transformation of Rat-1 cells by Ret is reversible. AP-stimulated macroscopic colonies formed in soft agar were harvested and further expanded under normal culture conditions in the presence or absence of AP stimulation. Three weeks after the removal of AP, cells had re-adhered strongly to the culture plates and displayed normal contact inhibition, whereas these two events were not seen in the continuous presence of AP (data not shown). As expected, Ret activation was no longer detectable in the absence of AP and these cells could no longer grow in soft agar (Fig. 1, bottom). Therefore, Rat-1 cell transformation induced by controlled homodimerization of Ret is reversible.


Figure 1
View larger version (17K):
[in this window]
[in a new window]
  		Figure 1. Reversible Ret-dependent transformation of Rat-1 cells. Top, cells expressing Ret-Fv were unstimulated or stimulated with AP as indicated. Ret-Fv- or MEN2A-expressing cells were assayed for growth in soft agar medium. Columns, mean values of the number of macroscopic colonies from a representative experiment of four; bars, SD. Middle, harvested colonies were further expanded with or without AP stimulation and the level of Ret activation was monitored by immunoblotting with an anti-phospho-Ret. The doublet corresponds to the mature and immature form of Ret (12). Immunoblotting with an anti-Ret (Total Ret) served as loading control. Bottom, these expanded cells were further assayed for growth in soft agar as described above. Columns, mean values from a representative experiment of two; bars, SD.

 
Reversible Rat-1 cell transformation correlates with alterations of DRM-associated cell surface proteins. Next, we explored potential transformation-related alterations of cell surface proteins that are associated with DRMs. Cell surface proteins from nontransformed Rat-1 cells were labeled with biotin and DRMs were isolated using a flotation assay based on resistance to solubilization by Triton X-100 at 4°C and buoyancy at low-density fractions of a bottom-loaded discontinuous sucrose gradient (Fig. 2A ; ref. 13). DRMs were recovered in fractions 1 and 2, which were highly enriched in flotillin, a proposed resident of membrane rafts (15). Less than 1% of the total biotinylated cell surface proteins were recovered in the DRM fractions of Rat-1 cells. Pooled DRM fractions from nontransformed cells were then compared with pooled DRM fractions from cells that were transformed either by Ret MEN2A or by AP-stimulated Ret-Fv. DRMs from transformed cells showed a significant alteration in the expression of biotinylated proteins (Fig. 2B), with an increase in three bands of ~80, 60, and 30 kDa. Indeed, the increase in expression of these bands paralleled the transformation process, as reversion of this phenotype by removal of AP, resulted in the decrease in expression of these bands. The variation in expression seems to be specific, as the expression level of flotillin was not affected during cellular transformation. These data show a correlation between Ret-induced transformation of Rat-1 cells and alterations of the DRM-associated cell surface proteins.


Figure 2
View larger version (46K):
[in this window]
[in a new window]
  		Figure 2. Alteration of DRM-associated cell surface proteins during transformation. A, DRMs from cell surface biotin-labeled Rat-1 cells were separated on a bottom-loaded sucrose step gradient. Probing the membrane with streptavidin-HRP or with antibody against flotillin assessed the distribution of proteins within fractions. Molecular size standards (in kDa) are shown at the right of the panel. B, pooled DRMs from nontransformed cells [WT Ret–expressing (WT) or nonstimulated Ret-Fv–expressing (none) cells] were compared with DRMs from transformed cells (MEN2A- or AP-stimulated RET-Fv–expressing cells) and transformation-reverted cells (revert, as defined in Fig. 1). Biotinylated cell surface proteins were revealed with streptavidin-HRP. Immunoblotting with antiflotillin served as a loading control. The position of molecular weight markers (in kDa) is shown on the right of the panel. Arrows, bands corresponding to p80, p60, and p30.

 
Standard CD44 is a major DRM-associated protein whose expression is up-regulated during Rat-1 cells transformation. To identify protein components of the bands described above, DRMs from Ret-transformed cells were isolated. Purified DRMs were further solubilized to prevent lipid-mediated protein/protein association. Biotinylated proteins were then extracted by avidin-coupled bead precipitation and resolved onto SDS-PAGE gels. Coomasie-stained bands were excised from the gels and subjected to proteomic analysis using LC-MS/MS (see Materials and Methods). Nine peptides were identified from the trypsin digested 80 kDa band. All corresponded to CD44 (see relevant sequences in Fig. 3A ), suggesting that this protein is a major component of the band. Surprisingly, in the gel-excised 30 kDa band, 8 of 13 peptides corresponding to the cytoplasmic tail of CD44 were detected. This probably does not reflect a degradation of the preparative protein samples, as we identified several peptides from proteins of the expected molecular weight in gel-excised bands, including the transmembrane proteins aminopeptidase A and the multispan MDR protein.5 Rather, these results suggest a shedding of the ectodomain of CD44 that generates membrane-tethered fragments (16).


Figure 3
View larger version (52K):
[in this window]
[in a new window]
  		Figure 3. Expression of standard CD44 form is increased during transformation. A, sequence coverage of amino acid sequence of rat CD44 by peptides identified by LC-MS/MS is represented by lines spanning the corresponding part of the sequence. The transmembrane domain of CD44 is boxed. B, pooled DRMs from nontransformed (none or WT), transformed (AP or MEN2A), and transformation-reverted (revert) cells, as described in Fig. 2, were immunoblotted with anti-CD44 antibody. Antiflotillin served as a loading control. Numbers, different Ret-Fv and MEN2A cellular clones. C, mRNA expression of CD44 in Rat-1 cellular clones expressing either WT Ret (WT) or MEN2A mutants. Numbers, different clones. RT-PCR products were visualized by ethidium bromide staining.

 
To ascertain that CD44 is indeed a component of the 80 kDa increased band during transformation, DRM-associated proteins from nontransformed and transformed cells were immunoblotted with specific CD44 antibodies. Figure 3B shows an increased intensity of CD44 staining in transformed cells (both AP and MEN2A) compared with nontransformed cells. Moreover, reversion of the transformation phenotype by removal of AP led to a decrease in CD44 expression. Thus, we confirmed a positive correlation between the increased expression of CD44 and transformation of Rat-1 cells.

We then investigated which of the CD44 isoforms were increased (1). CD44 was cloned from Ret WT or Ret MEN2A expressing clones by reverse transcription-PCR (RT-PCR) using primers that could identify both standard and variant CD44 isoforms. RT-PCR products showed a main band corresponding to the 1,191 bp standard form of CD44 (Fig. 3C). Sequence analysis of the PCR products confirmed the absence of variant isoforms of CD44 in the clones tested (data not shown). Altogether, these data show that standard CD44 form is a DRM-associated protein that increases in expression during the Ret-mediated transformation process of Rat-1 cells.

Transformed Rat-1 cells display an increased expression of CD44-ICD. Our initial hypothesis was that hyaluronan binding to CD44 might promote Ret-induced transformation of Rat-1 cells. We tested this by treating MEN2A-transformed cells with bovine testis hyaluronidase, which disrupts extracellular hyaluronan-cell interactions (17). This treatment did not affect the anchorage-independent growth capability of the cells (data not shown). Next, we turned our attention to a potential role for CD44 cleavage, as suggested by the mass spectrometric data. To detect CD44-ICD, we generated a COOH terminus GFP-tagged CD44-FL and a COOH terminus GFP-tagged CD44 cytoplasmic tail (CD44-ICD; refs. 7, 9; Fig. 4A ). CD44-FL stably expressed in Ret WT (WT44) or Ret MEN2A (MEN2A44) clones, was revealed by anti-GFP antibody staining as a main 105 kDa band with two additional fainter bands of ~45 and ~38 kDa (Fig. 4B). The main 105 kDa band corresponds to CD44-FL. Interestingly, expression of the 45 kDa band was predominant in nontransformed WT44 cells, whereas both the 45 and 38 kDa bands were observed at a similar intensity in transformed MEN2A44 cells.


Figure 4
View larger version (34K):
[in this window]
[in a new window]
  		Figure 4. Increased expression of CD44-ICD in Ret-transformed cells. A, schematic representation of GFP-tagged CD44 constructs used. Star 1, cleavage site by metalloprotease; star 2, cleavage site by {gamma}-secretase. The corresponding inhibitory drugs are indicated. The starting amino acid for CD44-ICD is shown. B, cellular lysates from clones stably expressing CD44-FL (WT44 or MEN2A44) or CD44-ICD were prepared and CD44 was analyzed by immunoblotting with an anti-GFP antibody. The position of CD44-FL and CD44-ICD are shown on the right of the panel, and molecular-size standards (in kDa) are shown on the left of the panel. C, WT44 or MEN2A44 clones were incubated with protease inhibitors as indicated (see text for details) and CD44 products were revealed by immunoblotting with anti-GFP antibody. The position of the 40 kDa molecular weight marker is shown on the left. D, WT44 clone were treated with hyaluronidase (HAase) in the presence or absence of protease inhibitors as indicated. The 45 and 38 kDa CD44 products were revealed by immunobloting with anti-GFP antibody. Immunoblotting with anti-Erk served as a loading control.

 
To clarify the nature of these bands, we made use of the broad spectrum metalloprotease inhibitor BB94, as metalloproteases can trigger cleavage of the CD44 ectodomain, leaving CD44-CTF (4). These fragments can be further cleaved by {gamma}-secretase into released CD44-ICD (9). BB94 treatment of WT44 cells strongly reduced the expression of the 45 kDa band, indicating that it is a CD44 cleavage product (Fig. 4C). In contrast, treatment of these cells with {gamma}-secretase inhibitors DAPT and L-685,458 (8), resulted in an increase in the expression of the 45 kDa plus an additional fainter band of ~55 kDa band. The latter likely reflects the existence of multiple cleavage sites in the membrane-proximal region of CD44 (18). These results are coherent with the accumulation of CD44-CTF fragments that would no longer be processed into CD44-ICD. Similar results were obtained in MEN2A44 cells. Therefore, the 45 kDa band would seem to be CD44-CTF.

Surprisingly, treatment of RET MEN2A cells with {gamma}-secretase inhibitors did not markedly alter the expression of the 38 kDa, which remained high (Fig. 4C). To gain further insights into the nature of this band, CD44-ICD was expressed in Ret WT clones. It migrated close to the 38 kDa band detected in MEN2A44 cells (Fig. 4B). The slight difference in the size of the two bands may reflect posttranslational modifications of the endogenously generated CD44 cleavage product, such as phosphorylations (19). It is of note that expression of this 38 kDa band, but not of CD44-CTF, was impaired in cells expressing CD44 mutants lacking two cysteine residues (see Fig. 6C), indicating that it is a CD44 product. Thus, these results suggest that the 38 kDa band is CD44-ICD. Next, we treated WT44 cells with hyaluronidase. Hyaluronidase is known to generate hyaluronan oligosaccharides (17), which, in turn, can promote CD44 cleavage (20). Hyaluronidase treatment resulted in an increased expression of both CD44-CTF and the 38 kDa band (Fig. 4D). This increased expression was strongly reduced in cells that had been preincubated with BB94. In addition, preincubation with DAPT resulted in an increase in the expression of CD44-CTF while the expression of the 38 kDa band was inhibited. Therefore, the 38 kDa band would seem to be CD44-ICD as its expression requires the sequential metalloprotease and {gamma}-secretase cleavage of CD44.


Figure 6
View larger version (25K):
[in this window]
[in a new window]
  		Figure 6. Release of CD44-ICD may relate to an association with DRMs. A, schematic representation of CD44-FL-C. Arrows, mutated cysteine residues. B, DRMs from clones stably expressing CD44-FL (WT44) or CD44-FL-C (WT44C–) were separated as in Fig. 2. The distribution of CD44 within fractions was detected by immunoblotting with anti-GFP antibody. Quantification of CD44 within fractions was assessed by densitometry analysis and is indicated below each blot. Immunoblotting with antiflotillin identifies the DRM fractions. C, cellular lysates from MEN2A clones stably expressing CD44-FL (MEN2A44) or CD44-FL-C (MEN2A44C–) were prepared and CD44 was analyzed by immunoblotting with an anti-GFP antibody as indicated. The position of CD44-CTF and CD44-ICD are shown on the right of the panel. Immunoblotting with anti-Erk served as a loading control.

 
Altogether, these results indicate that CD44 undergoes a sequential cleavage in Rat-1 cells with an increased expression of CD44-ICD in transformed cells.

CD44-ICD participates in the Rat-1 transformation process. We next explored the relationship between CD44-ICD expression and Rat-1 cell transformation. CD44-ICD was stably expressed in nontransformed Rat-1 cells. Consistent with previous reports (9), GFP-tagged CD44-ICD translocated into the nucleus of these cells (Fig. 5A ). Moreover, these cells acquired the capacity to grow in soft agar, whereas control cells expressing either empty vector or CD44-FL did not (Fig. 5B). However, in contrast to MEN2A-induced transformation of Rat-1 cells, CD44-ICD was not sufficient to fully abolish contact inhibition (data not shown). Thus, it seems that CD44-ICD can partially transform Rat-1 cells, strongly suggesting that it is part of the Ret MEN2A-induced transformation process. To further substantiate the above finding, we evaluated the effect of {gamma}-secretase inhibitors on the anchorage-independent growth capability of Ret-transformed cells. DAPT treatment significantly reduced the number of both Ret MEN2A- and AP-stimulated Ret-Fv-induced macroscopic colonies formed in soft agar (Fig. 5C). Moreover, this partial inhibition (~37% on average) was specific of an effect on Ret-dependent transformation of Rat-1 cells because the drugs did not affect the growth in soft agar of CD44-ICD-expressing cells. Thus, altogether, these results show that the release of CD44-ICD participates in Ret-dependent transformation of Rat-1 cells.


Figure 5
View larger version (21K):
[in this window]
[in a new window]
  		Figure 5. CD44-ICD participates in Ret-induced transformation. A, images obtained using confocal microscope indicate a nuclear localization of CD44-ICD in cellular clone 6 (GFP). BF, bright field microscopy. Bars, 10 µm. B, cellular clones expressing none, empty vector (GFP), CD44-FL (FL), or CD44-ICD (ICD, two different clones are shown) were assayed for anchorage-independent growth in soft agar as described in Materials and Methods. Columns, number of colonies; bars, SD. C, effect of {gamma}-secretase inhibitor DAPT on anchorage-independent growth of cells transformed by MEN2A Ret mutants (two different clones), by AP-stimulated Ret-Fv or by CD44-ICD (two different clones). Columns, number of colonies; bars, SD.

 
Release of CD44-ICD may require an association of CD44 with DRMs. Acylation is thought to be one of the mechanisms for recruiting proteins into DRMs (21, 22) and CD44 is fatty acid acylated (23). To evaluate the functional relevance of DRM association to CD44 cleavage, we generated CD44-FL mutants (CD44-FL-C) where potential palmitoylation sites of the protein were mutated (C289A and C298A; Fig. 6A ). CD44-FL-C mutants stably expressed in Ret WT clones (WT44C–), showed normal cell surface expression (data not shown) but were significantly impaired in their capacity to associate with DRMs (compare WT44 versus WT44C–, Fig. 6B). It is of note that the cysteine mutations did not completely abolish CD44 association to DRMs (~40% reduction), indicating that they are not solely responsible for such an association. Interestingly, CD44-FL-C mutants stably expressed in MEN2A clones (MEN2A44C–) were no longer susceptible to {gamma}-secretase-mediated cleavage as indicated by the absence of CD44-ICD product, although they remained sensitive to the initial metalloprotease-mediated cleavage, as indicated by the presence of CD44-CTF (Fig. 6C). This is in contrast to CD44-FL (MEN2A44), which is cleaved into CD44-CTF and CD44-ICD fragments (see also Fig. 4). Therefore, these results indicate that both cysteine C289 and C298 are necessary for the {gamma}-secretase-dependent cleavage of CD44-CTF. In addition, they suggest a role for the specific location of CD44 in a DRM environment in this process.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The metalloprotease-dependent shedding of CD44, which releases the ectodomain of the molecule from the membrane-bound COOH-terminal fragment, is frequently observed in human tumors (24). This shedding is thought to play an important role in promoting the metastatic behavior of CD44-expressing tumor cells by the interaction of CD44 with hyaluronan and other extracellular matrix components (4, 25). The data we report here suggest that the shedding of CD44 may not only modulate tumor cell adhesion, but also affect tumor malignancy through the subsequent regulated intramembrane proteolysis–dependent release of CD44-ICD that possesses transformation-promoting activity.

We have previously observed that the oncogenic conversion of immortalized rodent fibroblastic cells, induced by a controlled dimerization of the tyrosine kinase receptor Ret that mimics MEN2A oncogenic activation (12), could be reverted upon removal of the dimerizing agent. The reverted transformed phenotype was confirmed by the inability of these cells to sustain growth in an anchorage-independent manner (see Fig. 1), an operational test of malignancy (26). In this study, this model was used to identify molecules whose expression correlated with this oncogenic conversion. As a preliminary screen, we did a proteomic analysis of molecules that associated with DRMs. Whether DRMs are related to lipid rafts is currently a matter of debate (27). Nonetheless, DRM extraction remains a useful tool to select certain membrane proteins that have a differential affinity for these detergent-resistant structures.

We showed that the expression level of biotin-labeled cell surface proteins that associated with DRMs varied during the Ret-induced transformation process of Rat-1 cells. Using spectrometric analysis and Western blotting, standard CD44 was identified as one of the molecules whose increased expression correlates with the transformed phenotype of Rat-1 cells. This finding was particularly interesting because the altered expression or dysfunction of CD44 had been shown to contribute to numerous pathologic conditions (1). Typically, the interaction of CD44 with its main ligand hyaluronan initiates signaling that contributes to the development of certain tumors (see ref. 2 for a review). However, this mechanism was not involved in the transformation of Rat-1 cells induced by Ret. Instead, our results suggest a novel role for CD44-ICD in this process. This is based on several lines of experimental evidence, as follows: (a) the expression level of CD44-ICD is increased in transformed cells; (b) the CD44-ICD fragment possesses a transforming activity; and (c) drug-mediated inhibition of the {gamma}-secretase activity (responsible for CD44-ICD release) results in a reduced capacity of Ret-transformed cells to grow in an anchorage-independent manner. The partial nature of this inhibition suggests that CD44-ICD cooperates with other Ret-dependent signaling pathways to promote full oncogenic transformation of Rat-1 cells.

The observed increase in expression of CD44-ICD in transformed cells compared with nontransformed cells was not simply a consequence of an increase in the expression level of CD44-FL, as CD44-ICD was barely detected in nontransformed cells that expressed comparable amounts of CD44-FL (see Fig. 4). This is in contrast to the generation of CD44-CTF, the amount of which correlated with the abundance of CD44-FL (data not shown). These results suggest that the {gamma}-secretase-dependent cleavage of CD44 is enhanced in transformed cells, and/or the stability of CD44-ICD is increased. In support of the latter, CD44-ICD expression was still detectable up to 24 hours after {gamma}-secretase inhibition, which, in turn, suggests a deregulation of CD44-ICD turnover. In this scenario, increased expression of CD44-FL might still be required to generate sufficient levels of CD44-ICD to regulate transcription.

These present findings were obtained in a transformation model of rodent cells. The process by which these cells acquire cancer cell growth characteristics in vitro is well defined and has yielded insights into the mechanistic details of tumor progression in humans. Nonetheless, rodent and human cells have distinct requirements for cellular transformation (28). Therefore, the contribution of CD44-ICD to the conversion of normal human cells to a tumorigenic state is currently under investigation. In this regard, whether there is an increase in CD44 expression that could participate in the promotion of MEN2A tumors, i.e., medullary thyroid carcinoma (29), is not currently known. An osteopontin-CD44v6 autocrine loop that sustains proliferation and invasiveness of thyroid follicular cells transformed by Ret/PTC oncogenes (generated by rearrangements of Ret) has been reported (30). It is worth noting that this osteopontin loop is unlikely to be involved in our cellular model because we have only detected standard CD44 form, which does not bind efficiently to osteopontin (1). Nonetheless, Ret activation, either by MEN2A mutations or rearrangements of the gene, may affect the expression of CD44 (standard form or isoforms) in different types of tumors. Indeed, the Ras/extracellular signal-regulated kinase pathway that can trigger an increase in expression of CD44 (31) is activated by both MEN2A mutants and Ret/PTC (29).

The current view in the literature suggests that CD44 cleavage plays an important role in promoting metastasis but not in the initiation of tumorigenesis (1). However, there is a report describing CD44 cleavage in human tumors (as detected by the presence of CD44-CTF), irrespective of the grade of malignancy in the tumor samples examined (6). As noted by the authors, their data suggests that CD44 cleavage may also affect the early stages of tumor development. Clearly, our results add support to such an hypothesis and suggest a general role for CD44-ICD in the promotion of different types of CD44-expressing tumors. Remarkably, treatment of Rat-1 cells with hyaluronidase resulted in CD44-ICD generation. This may represent one example of how CD44-ICD generation can be initiated, as levels of hyaluronidases are often increased in malignant tumors (2).

CD44 is known to interact with DRMs (32) via an interaction with its transmembrane domain (33). It was recently shown that the biologically active {gamma}-secretase complex also resides within DRMs (34). Therefore, the differential affinity of both of these proteins for detergent-resistant structures suggests that they may share a preference for a certain lipid environment in intact cells. In an attempt to address a functional relevance for the association of CD44 with DRMs, two potential palmitoylation sites of the protein were mutated. Mutated CD44 was not cleaved by {gamma}-secretase; however, the mutated protein remained susceptible to metalloprotease cleavage, as indicated by the detection of CD44-CTF but not CD44-ICD in MEN2A clones. This may be due to a conformational change in the transmembrane domain of CD44 whereby it is no longer accessible for cleavage by {gamma}-secretase. Alternatively, these results could support the need for a specific lipid environment for CD44 cleavage, as the mutated CD44 molecules displayed a reduced association with DRMs.

In conclusion, the Ret-induced Rat-1 cellular transformation model we described here uncovers a transforming activity for CD44-ICD and suggests a complex role for CD44 in neoplastic transformation. Indeed, CD44 cleavage may not only play an important role in promoting metastasis, but it may also play a relevant role in the neoplastic transformation process. The examination of CD44-ICD activity in human tumors that express CD44 now awaits.


   Acknowledgments
 
Grant support: Centre National de la Recherche Scientifique institutional grants; Association pour la Recherche sur le Cancer (SM 4312), Fondation de France (2002 010 893), and Ligue Nationale Contre le Cancer Comite du Rhone; fellowships from the French Ministry of Research (L. Pelletier), Hospices Civils de Lyon/Centre National de la Recherche Scientifique (B. Frêche), and Fondation de France (C. Luquain); and National Health and Medical Research Council of Australia (D. Christiansen).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank ARIAD for the regulation kits used in our work (www.ariad.com/regulationkits), Dr. B. Mari (IPMC/CNRS UMR6097, Valbone Sophia-Antipolis, France) for BB94, and Drs. T. Renno and F. Solari for helpful discussion.


   Footnotes
 
5 L. Pelletier, unpublished data. Back

Received 10/26/05. Revised 11/28/05. Accepted 1/26/06.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Ponta H, Sherman L, Herrlich PA. CD44: from adhesion molecules to signalling regulators. Nat Rev Mol Cell Biol 2003;4:33–45.[CrossRef][Medline]
  2. Toole BP. Hyaluronan: from extracellular glue to pericellular cue. Nat Rev Cancer 2004;4:528–39.[CrossRef][Medline]
  3. Bourguignon LY. CD44-mediated oncogenic signaling and cytoskeleton activation during mammary tumor progression. J Mammary Gland Biol Neoplasia 2001;6:287–97.[CrossRef][Medline]
  4. Okamoto I, Kawano Y, Tsuiki H, et al. CD44 cleavage induced by a membrane-associated metalloprotease plays a critical role in tumor cell migration. Oncogene 1999;18:1435–46.[CrossRef][Medline]
  5. Nagano O, Murakami D, Hartmann D, et al. Cell-matrix interaction via CD44 is independently regulated by different metalloproteinases activated in response to extracellular Ca(2+) influx and PKC activation. J Cell Biol 2004;165:893–902.[Abstract/Free Full Text]
  6. Okamoto I, Tsuiki H, Kenyon LC, et al. Proteolytic cleavage of the CD44 adhesion molecule in multiple human tumors. Am J Pathol 2002;160:441–7.[Abstract/Free Full Text]
  7. Lammich S, Okochi M, Takeda M, et al. Presenilin-dependent intramembrane proteolysis of CD44 leads to the liberation of its intracellular domain and the secretion of an Aß-like peptide. J Biol Chem 2002;277:44754–9.[Abstract/Free Full Text]
  8. Murakami D, Okamoto I, Nagano O, et al. Presenilin-dependent {gamma}-secretase activity mediates the intramembranous cleavage of CD44. Oncogene 2003;22:1511–6.[CrossRef][Medline]
  9. Okamoto I, Kawano Y, Murakami D, et al. Proteolytic release of CD44 intracellular domain and its role in the CD44 signaling pathway. J Cell Biol 2001;155:755–62.[Abstract/Free Full Text]
  10. Brown MS, Ye J, Rawson RB, Goldstein JL. Regulated intramembrane proteolysis: a control mechanism conserved from bacteria to humans. Cell 2000;100:391–8.[CrossRef][Medline]
  11. Landman N, Kim TW. Got RIP? Presenilin-dependent intramembrane proteolysis in growth factor receptor signaling. Cytokine Growth Factor Rev 2004;15:337–51.[CrossRef][Medline]
  12. Frêche B, Guillaumot P, Charmetant J, et al. Inducible dimerization of RET reveals a specific AKT deregulation in oncogenic signalling. J Biol Chem 2005;280:36584–91.[Abstract/Free Full Text]
  13. Vidalain PO, Azocar O, Servet-Delprat C, Rabourdin-Combe C, Gerlier D, Manie S. CD40 signaling in human dendritic cells is initiated within membrane rafts. EMBO J 2000;19:3304–13.[CrossRef][Medline]
  14. Ferro M, Salvi D, Riviere-Rolland H, et al. Integral membrane proteins of the chloroplast envelope: identification and subcellular localization of new transporters. Proc Natl Acad Sci U S A 2002;99:11487–92.[Abstract/Free Full Text]
  15. Liu J, Deyoung SM, Zhang M, Dold LH, Saltiel AR. The stomatin/prohibitin/flotillin/HflK/C domain of flotillin-1 contains distinct sequences that direct plasma membrane localization and protein interactions in 3T3-L1 adipocytes. J Biol Chem 2005;280:16125–34.[Abstract/Free Full Text]
  16. Nakamura H, Suenaga N, Taniwaki K, et al. Constitutive and induced CD44 shedding by ADAM-like proteases and membrane-type 1 matrix metalloproteinase. Cancer Res 2004;64:876–82.[Abstract/Free Full Text]
  17. Ohno-Nakahara M, Honda K, Tanimoto K, et al. Induction of CD44 and MMP expression by hyaluronidase treatment of articular chondrocytes. J Biochem (Tokyo) 2004;135:567–75.[Abstract/Free Full Text]
  18. Okamoto I, Kawano Y, Matsumoto M, et al. Regulated CD44 cleavage under the control of protein kinase C, calcium influx, and the Rho family of small G proteins. J Biol Chem 1999;274:25525–34.[Abstract/Free Full Text]
  19. Legg JW, Lewis CA, Parsons M, Ng T, Isacke CM. A novel PKC-regulated mechanism controls CD44 ezrin association and directional cell motility. Nat Cell Biol 2002;4:399–407.[CrossRef][Medline]
  20. Sugahara KN, Murai T, Nishinakamura H, Kawashima H, Saya H, Miyasaka M. Hyaluronan oligosaccharides induce CD44 cleavage and promote cell migration in CD44-expressing tumor cells. J Biol Chem 2003;278:32259–65.[Abstract/Free Full Text]
  21. Resh MD. Fatty acylation of proteins: new insights into membrane targeting of myristoylated and palmitoylated proteins. Biochim Biophys Acta 1999;1451:1–16.[Medline]
  22. McCabe JB, Berthiaume LG. N-terminal protein acylation confers localization to cholesterol, sphingolipid-enriched membranes but not to lipid rafts/caveolae. Mol Biol Cell 2001;12:3601–17.[Abstract/Free Full Text]
  23. Bourguignon LY, Kalomiris EL, Lokeshwar VB. Acylation of the lymphoma transmembrane glycoprotein, GP85, may be required for GP85-ankyrin interaction. J Biol Chem 1991;266:11761–5.[Abstract/Free Full Text]
  24. Nagano O, Saya H. Mechanism and biological significance of CD44 cleavage. Cancer Sci 2004;95:930–5.[CrossRef][Medline]
  25. Goebeler M, Kaufmann D, Brocker EB, Klein CE. Migration of highly aggressive melanoma cells on hyaluronic acid is associated with functional changes, increased turnover and shedding of CD44 receptors. J Cell Sci 1996;109:1957–64.[Abstract]
  26. Assoian RK. Anchorage-dependent cell cycle progression. J Cell Biol 1997;136:1–4.[Free Full Text]
  27. Lichtenberg D, Goni FM, Heerklotz H. Detergent-resistant membranes should not be identified with membrane rafts. Trends Biochem Sci 2005;30:430–6.[CrossRef][Medline]
  28. Rangarajan A, Hong SJ, Gifford A, Weinberg RA. Species- and cell type-specific requirements for cellular transformation. Cancer Cell 2004;6:171–83.[CrossRef][Medline]
  29. Santoro M, Carlomagno F, Melillo RM, Fusco A. Dysfunction of the RET receptor in human cancer. Cell Mol Life Sci 2004;61:2954–64.[CrossRef][Medline]
  30. Castellone MD, Celetti A, Guarino V, et al. Autocrine stimulation by osteopontin plays a pivotal role in the expression of the mitogenic and invasive phenotype of RET/PTC-transformed thyroid cells. Oncogene 2004;23:2188–96.[CrossRef][Medline]
  31. Hofmann M, Rudy W, Gunthert U, et al. A link between ras and metastatic behavior of tumor cells: ras induces CD44 promoter activity and leads to low-level expression of metastasis-specific variants of CD44 in CREF cells. Cancer Res 1993;53:1516–21.[Abstract/Free Full Text]
  32. Neame SJ, Uff CR, Sheikh H, Wheatley SC, Isacke CM. CD44 exhibits a cell type dependent interaction with triton X-100 insoluble, lipid rich, plasma membrane domains. J Cell Sci 1995;108:3127–35.[Abstract]
  33. Perschl A, Lesley J, English N, Hyman R, Trowbridge IS. Transmembrane domain of CD44 is required for its detergent insolubility in fibroblasts. J Cell Sci 1995;108:1033–41.[Abstract]
  34. Vetrivel KS, Cheng H, Lin W, et al. Association of {gamma}-secretase with lipid rafts in post-Golgi and endosome membranes. J Biol Chem 2004;279:44945–54.[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pelletier, L.
Right arrow Articles by Manié, S. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pelletier, L.
Right arrow Articles by Manié, S. N.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online