a three-dimensional and temporo-spatial model to study … · [cancer research 61, 81–87, january...

[CANCER RESEARCH 61, 81–87, January 1, 2001]

Advances in Brief

A Three-Dimensional and Temporo-Spatial Model to Study Invasiveness of CancerCells by Heregulin and Prostaglandin E2

1

Liana Adam, Abhijit Mazumdar, Tushar Sharma, Terence R. Jones,2 and Rakesh Kumar3

Department of Molecular and Cellular Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030

Abstract

To study the temporal expression of motile structures and proteaseactivity during colon cancer cell invasion by heregulin-b1 (HRG) andprostaglandin E2 (PGE2), we have developed a three-dimensional spatialmodel system. HRG and PGE2 each induced the formation of well-organized, three-dimensional structures with empty spaces in the centerand stimulated the expression of urokinase plasminogen activator (uPA)with differential localization of membrane-bound uPA at the focal adhe-sion points and leading edges of the motile cells. A specific cyclooxygen-ase-2 inhibitor blocked the formation of three-dimensional luminal glan-dular structures induced by HRG but did not block those induced byPGE2. A specific antagonist of uPA receptor completely blocked theformation of these luminal glandular structures induced by PGE2 andHRG. These findings suggest that HRG-mediated increased invasivenessof colon cancer cells is augmented at least in part by induction of PGE2and uPA, and this augmentation may involve the formation of three-dimensional invasive structures via the uPA pathway. In addition, thethree-dimensional model system presented here may have a wider appli-cation to screen the effects of therapeutic compounds and biomolecules ondifferent spatial aspects of colonic biology, including cell growth, motility,invasion, survival, and apoptosis.

Introduction

Accumulating evidence suggests that colonic tumorigenesis is par-tially regulated by the growth factor-inducible form of COX-2,4 anenzyme responsible for the conversion of arachidonic acid to PGs (1).PGs appear to play a variety of roles in the gastrointestinal tract,including participation in physiological processes such as cell motilityand in pathological processes such as neoplasia (1–3). The primaryPG generated in colon cancer tissue appears to be PGE2. The COX-2isoform is inducible and found in very low levels in normal tissues butin greatly increased levels in inflamed tissues. In addition, increasedlevels of PGE2, COX-2 protein, and COX-2 mRNA have been re-ported in colorectal adenomas and carcinomas, but not in adjacentnormal-appearing mucosa, suggesting a potential link between intes-tinal COX-2 activity and tumorigenesis (4–6). Shenget al. (7)showed that the tumorigenic potential of cells expressing high levelsof COX-2 could be reversed by COX-2 inhibitors, suggesting thepotential neoplastic role of PGE2 in colon cancer. Using an APCknockout mouse model, Oshimaet al. (8) further confirmed a role of

COX-2 expression in colorectal tumorigenesis and showed dramaticreductions in the number and size of intestinal polyps by a specificCOX-2 inhibitor. Furthermore, ectopic overexpression of COX-2 canalso lead to inhibition of apoptosis in colon cells (9, 10) and increasedmetastatic potential of colon cancer (11). Alterations in cytoskeletonreorganization in head and neck cancers have been shown to berelated to PGE2-mediated modulation of cell adherence (12).

Recent studies suggest that in addition to the COX-2 pathway, theinvasiveness of colon cancer can be also influenced by growth factorssuch as HRG and serine proteases such as uPA. For example, HRG,which binds to the HER-3 and HER-4 receptor (13), regulates theprogression of cancer cells to a more invasive phenotype (14). Re-cently, we confirmed that in the absence of HER-2 (also known asc-erbB2 or c-neu) overexpression, HRG activation of breast and coloncancer cells promotes the development of a more invasive phenotypein these cells (14, 15). The mechanisms and pathways by which HRGinfluences the biology of colon cancer cells remain elusive.

One of the important steps for tumor progression and invasion is thedestruction of the ECM that separates the epithelial and stromalcompartments by serine proteinases, thus facilitating the migrationand intravasation of cancer cells (16). The uPA system-mediatedproteolysis contributes to the dissolution of connective tissues sur-rounding the cancer cells and of the perivascular basement membrane.Evidence suggests that the progression of colon cancer cells alsoinvolves the uPA, which has its own high-affinity cell surface recep-tor, uPAR, that greatly enhances the action of uPA on plasminogen(16–18). uPAR localizes on cell-cell junctions and toward the leadingedge of invading cells (19, 20). Regulated, spatially localized degra-dation of the ECM is required for tissue remodeling and invasiveness.Thus, uPA is spatially and metabolically positioned to play a pivotalrole in the directed cascade of protease activity required for tissueinvasion by the tumor cells.

The expression of uPA and uPAR have been demonstrated inessentially every solid tumor type examined to date (16). Expressionis not restricted to the tumor cells themselves; several tumor-associ-ated cell types including macrophages, mast cells, endothelial cells,natural killer cells, and fibroblasts are all capable of uPA and uPARexpression (16). The pattern of expression in these cells differs,depending on the type of tumor. For example, tumor cells that exclu-sively express uPAR (but not uPA), such as some forms of coloncancer, can recruit uPA made by surrounding stromal cells (20).Tumor cell invasiveness and aggressiveness correlate strongly withthe expression of uPAR. Expression of uPA and uPAR is oftenrestricted to the leading edge of tumor progression and the tumor-hostinterface (21, 22).

Our previous studies showed that HRG stimulates the secretion ofPGE2 andin vitro invasiveness of human colon cancer cells (15). It isnot known, however, whether uPA plays a role in the actions of HRGand PGE2 in colon cancer cells. The purpose of this study was toestablish whether HRG-mediated stimulation of colon cancer inva-siveness could be influenced by PGE2 and uPA. To demonstrate this,we used a three-dimensional model in which we could modulate the

Received 3/15/00; accepted 11/8/00.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby markedadvertisementin accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by NIH Grant CA80066, American Institute for Cancer ResearchGrant 96A077, and Cancer Center Core Grant CA16672.

2 Present address: Angstrom Pharmaceuticals, Inc., 11585 Sorrento Valley Road, SanDiego, CA 92121.

3 To whom requests for reprints should be addressed, at Department of Molecular andCellular Oncology, The University of Texas M. D. Anderson Cancer Center 108, 1515Holcombe Boulevard, Houston, TX 77030. Fax: (713) 745-3792; E-mail: [email protected].

4 The abbreviations used are: COX-2, cyclooxygenase-2; PG, prostaglandin; HRG,heregulin-b1; HER, human epidermal growth factor receptor; uPA, urokinase plasmino-gen activator; uPAR, uPA receptor; ECM, extracellular matrix; DMEM:F12, DMEM:Ham’s F-12; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide.

81

Research. on November 20, 2020. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

architectural remodeling of FET colon cells by blocking differentlevels in the proposed HRG-induced uPA activation pathway. Usinga simplified in vitro model, this study demonstrated for the first timethat both HRG and PGE2 augmentation involve the formation ofdistinct three-dimensional glandular structures. Because HRG is aparacrine growth factor, these findings raise the possibility that HRG-and PGE2-mediated up-regulation of uPA in tumor cells may enhancethe ability of colon tumor cells to degrade the extracellular milieu andthereby invade normal tissue.

Materials and Methods

Cell Culture and Reagents.FET human colon carcinoma cells (15, 23)were maintained in DMEM:F12 (1:1) supplemented with 10% FCS. Anti-epidermal growth factor receptor antibody and HRG were purchased fromNeomarkers, Inc. (Fremont, CA). Theb-catenin polyclonal antibody (24) wasa gift from P. McCrea (The University of Texas M. D. Anderson CancerCenter, Houston, TX). The anti-uPA monoclonal antibody and recombinantlow molecular weight uPA were purchased from American Diagnostica. TheCOX-2 inhibitor NS398 was purchased from Sigma.

Boyden Chamber Assay.The effect of HRG on the invasiveness of FETcolon cells was determined by the Boyden chamber assay (14). Briefly, thebottom of the porous 8mm filter was coated with a thick layer of Matrigel/DMEM:F12 (1:2) that serves as a chemoattractant for the cells that are platedon the upper side of the filter. After 12 h, the cells that pass through the poresand reattach on the lower surface of the filter were quantitated. The filters werefixed, stained with propidium iodide (0.1%, w/v), mounted on a slide, andanalyzed by confocal microscopy. The cells that invaded the Matrigel wereconsidered invasive, and results were expressed as the percentage of totalcells/microscopic field (340 magnification) from both sides of filter plus thoseen passagethat were inside the pores.

Formation of Lumen-containing Three-dimensional Structures. Cellswere allowed to grow on a thick layer of Matrigel. Briefly, 100ml of

Matrigel/DMEM:F12 (diluted 1:2) were added to each well of a Lab-Tekeight-chamber slide (Molecular Probes) and allowed to gel for 30 min at 37°C.FET cells were trypsinized, washed sequentially in DMEM:F12 supplementedwith 10% serum and DMEM:F12 without serum, and resuspended into singlecells, and 10,000 cells were resuspended in 400ml serum-free DMEM:F12/well. After 10 min, HRG (10–100 ng/ml) or PGE2 (10–100 ng/ml) was added,and slides were monitored for several days for morphological changes. Aconcentration of 30 ng/ml HRG and 50 ng/ml PGE2 was able to induceformation of cellular clusters after 48 h for HRG and after 8 h for PGE2. Insome cases, we used Matrigel/DMEM:F12 (diluted 1:4)-coated Lab-Tek cham-bers to analyze changes in cellular shapes reminiscent of a motile phenotype.All morphological changes were analyzed after immunostaining and confocalmicroscopy analysis.

Northern Analysis. Total cytoplasmic RNA was analyzed by Northernblot analysis using cDNA probes against uPA and uPAR (American TypeCulture Collection, Manassas, VA).

Growth Assay. Cell proliferation was measured by the MTT dye (Sigma)uptake method, as described previously (25). About 7000 cells/0.5 ml culturemedium were seeded into each well of a 24-well plate. After 24 h, appropriatecultures were supplemented with HRG or PGE2. For each time point, themedium was removed from triplicate wells, MTT (5 mg/ml PBS) was added,and the plate was wrapped in aluminum foil and kept at 37°C for 4 h. The dyesolution was aspirated, and the dye taken up by the cells was extracted in 1 mlof isopropyl alcohol:1N HCl (96:4) for determination of absorbance at 570 nm.

Immunostaining and Confocal Analysis. For immunofluorescence stain-ing, cells were grown on uncoated or Matrigel-coated Lab-Tek chamber slides.After appropriate treatments, cells were fixed for 10 min in a 1:2 mixture ofcold methanol and ethanol and incubated with the indicated primary antibodiesfollowed by the respective FITC- or rhodamine-conjugated secondary antibod-ies (Molecular Probes). Confocal imaging, three-dimensional analysis, andintensity quantification were performed using a Zeiss LSM and personalcomputer-based LSM510 Version 2.01 software (14, 15).

Fig. 1. HRG and PGE2 induce cytoskeleton modifi-cations in colon carcinoma cells. FET cells cultured on(A–C) plastic or (D–F) Matrigel were processed forindirect immunofluorescence microscopy. Representa-tive examples of rhodamine-phalloidin staining (A–C)show the distribution of F-actin; representative examplesof FITC-conjugated antimouse secondary antibody stain-ing (D–F) indicate the localization ofb-tubulin. G, theBoyden chamber assay showing the effects of HRG (30mg/ml, 12 h) and PGE2 (50 mg/ml, 12 h) on the invasionof FET cells.H, MTT assay showing the effects of HRG(30 mg/ml, 48 h) and PGE2 (50 mg/ml, 48 h) on thegrowth of FET cells.

82

A SPATIAL MODEL FOR COLON CELLS INVASIVENESS



Results and Discussion

HRG and PGE2 Induce Cytoskeleton Modifications. We haveshown previously that treatment of colon cancer cells with HRG leadsto increased production of PGE2 (15). Here we compared the effectsof HRG and PGE2 on the motility of a noninvasive FET colon cancercell line. Incubation of colon cells grown on an uncoated surface withHRG or PGE2 triggered cytoskeleton modifications in the FET cellsthat were visualized by F-actin staining (Fig. 1,A–C); these modifi-cations were reminiscent of those in cells acquiring a more motilephenotype. HRG and PGE2 treatments induced membrane ruffling(arrows) and changes in cell shape. In contrast to the uncoated slide,when cells were grown on Matrigel, the effect of PGE2 was differentfrom that of HRG, as assessed by staining withb-tubulin antibody(Fig. 1, D–F). HRG treatment induced the formation of oblong cellclusters (Fig. 1E); the PGE2 treatment induced the formation ofintracellular pseudoglandular structures containing large empty spaces(Fig. 1F). Consistent with the invasive nature of PGE2-treated cells

grown in Matrigel, PGE2 promoted the invasion of FET cells usingBoyden chamber assays (Fig. 1G) but did not have any effect on cellgrowth as measured by the MTT assay after 48 h of treatments (Fig.1H). In contrast, HRG induced cell invasion as well as cell prolifer-ation (Fig. 1,G andH, respectively).

Three-dimensional Structures Induced by PGE2 Differ fromThose Induced by HRG. Treatment with PGE2 of cells plated onMatrigel induced the formation of three-dimensional cellular struc-tures that were visible by phase-contrast microscopy after 12 h oftreatment. The treatment of FET cells with HRG induced the forma-tion of three-dimensional structures that were visible after 48 h oftreatment (Fig. 2A). The results of phase-contrast microscopy studiessuggested that the structures induced by PGE2 were morphologicallydifferent from those induced by HRG and characterized by differenttime kinetics. Using specific markers such asb-catenin for intercel-lular junctions or propidium iodide for nuclear DNA, we analyzedthese structures by confocal scanning microscopy followed by three-

Fig. 2. A, phase-contrast photomicrographs of RG- or PGE2-treated FET cells plated on Matrigel.B, b-catenin immunostaining and confocal microscope analysis followed bythree-dimensional reconstruction. An overall view is shown in thetop left panel. Thebottom left panelshows a lateral view of the reconstructed structure. Two of its virtual confocalsections taken at different levels (red arrowandblue arrow) are also shown in thetwo right panelsto emphasize the existence of multilumens, (yellow arrows).C, b-catenin andpropidium iodide costaining and confocal analysis with three-dimensional reconstruction show the monoluminal, three-dimensional structures formed after PGE2 treatment of FET cells.An overall three-dimensional view from the top (top left panel) and a lateral three-dimensional view (top right panel) are shown. Vertical section,red arrow; horizontal sections,bluearrows.D, overall three-dimensional view (left panels) with the corresponding virtual optical section (right panels) showing the lack of lumen-containing structures in FET cells treatedwith transforming growth factor-a.

83




dimensional reconstruction technique. Our results showed that after48 h of HRG treatment, the cells displayed intracellular multiluminalthree-dimensional structures that appeared to be interconnected atdifferent levels (Fig. 2B, yellow arrows). Two horizontal sectionsshowing the apical (red arrow) and basal (blue arrow) levels of thestructure are also shown (Fig. 2B, right panels). In contrast to HRGtreatment, the three-dimensional structures in the PGE2-treated cellswere monoluminal (Fig. 2C), oriented perpendicularly on the slidesurface, often twice as long as the HRG-induced structures, andresembled colonic glandular-like structures (Fig. 2C). Three differentsections were virtually cut at the basal level (yellow arrow pointing toboth sides), at the apical level (yellow arrow pointing to the left), andvertically in the Y axis (red arrow; Fig. 2C). These structures inPGE2-treated cells were in a state of continuous remodeling, asindicated by the nuclear DNA staining, which showed mitotic cells(white arrow) at the basal level and apoptotic cells at the apical level(white arrowhead; Fig. 2C). Interestingly, no lumen-containing three-dimensional structures were induced within the cells when transform-ing growth factora was used as a signal (Fig. 2D). These findingssuggest that the formation of three-dimensional invasive structuresrequires ligand-triggered signals (Fig. 2D).

HRG and PGE2 Differentially Modulate the Expression Leveland Temporal Distribution of uPA. Modification of the invasivepotential of a cell is generally accompanied by rearrangement of thecytoskeleton and by changes in the pericellular proteolytic activity, wherethe uPA/uPAR system is playing an important role. Recently, we showedthat HRG up-regulates the expression and functions of uPA/uPAR inbreast cancer cells (26). In the present study, we analyzed whether eitherHRG or PGE2 up-regulates the expression of uPA and uPAR mRNA inFET cells (26). Northern blot analysis demonstrated that HRG (30 ng/ml), but not PGE2 (50mg/ml), up-regulated the level of uPAR mRNA tolevels higher than those seen in untreated control cells (Fig. 3A). BothHRG and PGE2 up-regulated the levels of uPA mRNA, although they didso to different extents (Fig. 3A).

Because spatially localized degradation of the ECM is important fortissue remodeling and directed motility and invasion, we used confo-cal microscopy and three-dimensional reconstruction methods to an-alyze the temporal distribution of uPA bound to the plasma membraneof the cells. First, we examined the effect of HRG (30mg/ml, 12 h)and PGE2 (50 ng/ml, 12 h) on the distribution of uPA in FET cellsplated on a thin layer of Matrigel (Fig. 3B). Both HRG and PGE2

treatments substantially altered the cell shapes compared with notreatment in the control cells. These morphological modificationswere accompanied by a dramatic redistribution of uPA, both at thelevel of focal adhesion points and, in the case of PGE2 treatment, atthe site of dome-like structures (Fig. 3B,lower andupper, respective-ly). More interestingly, the uPA redistribution occurred most fre-quently at the leading edges of the cells, followed in incidence byredistribution at the focal adhesion-like complexes in the case of cellstreated with HRG. A low, uniform, peripheral localization of uPAcould be observed in the case of the untreated FET cells (Fig. 3B). Theepidermal growth factor receptor was costained to identify the cellmembranes, including the free edges of FET cells, because costainingwith b-catenin would identify mainly the intercellular junctions.

To evaluate the status of uPA distribution in the three-dimensionalclusters induced after 48 h of PGE2 and HRG treatment, cells werecostained forb-catenin and uPA proteins and analyzed using confocalmicroscopy and three-dimensional reconstruction (Fig. 4). We alsoquantified the intensities of the pixels in each channel on the pre-defined horizontal sections, where the values are plotted on (a) a redhistogram for the Alexa546 fluorochrome channel that corresponds tob-catenin and (b) a green histogram for the FITC channel that corre-sponds to uPA. Superimposition of thered andgreen peakssuggeststhe colocalization of proteinsb-catenin and uPA (Fig. 4). In thestructures induced by HRG, uPA was distributed mainly on theinternal membrane facing the empty space of the structure and occa-sionally on cell membrane projections, which, as an example, arerepresented by themiddle peakshowing high pixel intensities in boththe red andgreen channels. This was different from the uPA expres-sion pattern in the PGE2-induced structures, in which uPA was locatedon the cell membrane facing the outside surface of the cellular walls(Fig. 4). The levels of membrane-bound uPA were also different.Thus, the staining intensity of uPA in HRG-treated cells was two tothree times higher than that in the PGE2-treated cells. However, PGE2

treatment was accompanied by an increase in the level of uPA stainingcompared with that found in the untreated control cells. In brief, theseresults suggested that the temporal uPA distribution and intensities inthree-dimensional structures were differentially influenced by HRGversusPGE2. However, both HRG and PGE2 increased the level ofmembrane-bound uPA above the levels found in untreated controlcells.

uPAR/uPA Interaction Is Necessary for the Formation of Lu-men-containing Three-dimensional Structures.To understandhow the three-dimensional structures are formed, we analyzed uPA

Fig. 3. HRG and PGE2 differentially modulate the expression level and temporaldistribution of uPA.A, Northern blot analysis of total mRNA for the expression of uPAR,uPA, and GAPDH.B, FITC-conjugated antimouse antibody staining of FET cells platedon a thin layer of Matrigel, treated with HRG or PGE2, and processed for indirectimmunofluorescence microscopy shows the distribution of uPA; Alexa546-conjugatedantirabbit secondary antibody indicates the localization of epidermal growth factor recep-tor as a membrane marker. Sections were taken at two different levels (upperand lower)in PGE2-treated cells.

84




Fig. 4. Quantification of membrane-bound uPA in FET cells plated on Matrigel.Cells were costained forb-catenin and uPA and analyzed by confocal microscopy,followed by quantification of pixel intensities in each channel on predefined confocalZ sections. Numerical values are plotted on ared histogramfor the channel corre-sponding tob-catenin and on agreen histogramfor the channel corresponding touPA. Three different sections are shown for untreated cells (CON), HRG (30 ng/ml,48 h)-treated cells, or PGE2 (50 ng/ml, 48 h)-treated cells.

85




distribution during the very initial phases of cellular movement on athick layer of Matrigel in response to a 12-h treatment with HRG anda 6-h treatment with PGE2 (Fig. 5,A andB). Thus, we could identifythe appearance of “hot spots” (arrows) on the cell-cluster surface thatare not in direct contact with the Matrigel; these hot spots are char-acterized by uPA-bound andb-catenin-positive membrane projec-tions. They represent sites where the cellular junctions are discon-nected, and the resulting free edges of the cells are flipped to allow theinitiation of lumen formation (Fig. 5A, arrows). The accumulation ofmembrane-bound uPA (hot spots) appeared to be an early event thatcorrelated with the presence of luminal three-dimensional structures.

To understand the potential contribution of COX-2 and uPA path-ways in the cellular remodeling leading to the formation of luminalthree-dimensional structures (Fig. 5,arrows), we examined the effectsof a specific COX-2 inhibitor, NS398 (15), and a specific uPARantagonist, Å36, which acts as a competitive inhibitor (IC50 5 5 nM)of uPAR on colon cancer RKO cells (27). Interestingly, cotreatmentof cultures with both HRG and Å36 blocked the formation of lumencontaining three-dimensional clusters (Fig. 5C). Cotreatment of cellswith HRG and COX-2 inhibitor NS398 also blocked the formation ofa lumen inside the three-dimensional clusters as well as the HRGeffect on uPA distribution, despite a slight increase in uPA staining atthe level of the focal adhesion complexes (Fig. 5C). Treatment of thecells with NS398 did not affect the formation of PGE2-inducedthree-dimensional structures (Fig. 5C). Treatment of the cells withNS398 blocked the HRG-triggered increase in the invasiveness ofFET cells (Fig. 5D). Similarly, treatment with Å36 inhibited the HRG-and PGE2-triggered invasiveness of FET cells (Fig. 5D). These resultssuggest that PGE2 synthesis may be necessary for the formation ofHRG-induced cell structures on Matrigel and that both PGE2 and

HRG may use the uPA pathway for the formation of luminal three-dimensional structures. These events may be closely linked to theinvasiveness of colon cancer cells. Colon cell lines are known todevelop a distinct crypt-like architecture when cocultured with fetalrat mesenchyme cells (28) or implanted under the kidney capsules ofSwissnu/nu(nude) mice and allowed to grow for 10 days (29). Thesemodels, although very laborious, emphasized the need for a veryactive cellular stroma that originates from the animal host and leads todifferent types of glandular structures. Substitution of the stroma withexogenously added HRG in our model enabled us to study, in a moreaccessible way, distinct multicell architectural modifications, as wellas dynamics of different constituents such as cytoskeletal proteins orproteolytic enzymes.

In summary, the results presented here suggest that HRG affectscolonic tissue remodeling, at least, in part, by PGE2 up-regulation. Inthis model, HRG can induce a more invasive phenotype, probably dueto its effects on the uPA/uPAR pathway as well as up-regulation ofPGE2, which can, in turn, also feed into the uPA pathway. Althoughboth HRG and PGE2 can trigger a series of distinct morphogeneticalterations, PGE2 was more potent that HRG. It is possible that overa period of time, the HRG-induced PGE2 may attain a significantlyhigher level in the conditioned medium. In addition, there may beother unidentified effects of HRG that could partially attenuate PGE2-stimulated remodeling, and HRG-treated cells may have a greateramount of bound uPA than PGE2-treated cells. Additional studies arerequired to address these and other possibilities. There is a regulatoryinterplay between the stroma-secreted ECM proteins and the epithelialcells that harbor various surface receptors. This interplay may becritical for a dynamic control of cellular movement and cytoskeletonorganization. HRG and PGE2 heavily modulate these events in the

Fig. 5. Effects of specific inhibitors on the formation of three-dimensional structures by HRG and PGE2. FET cells plated on Matrigelwere treated with various combinations of PGE2, HRG, NS398, andÅ36 and processed for indirect immunofluorescence microscopy.A andB, whole-image projection show early morphogenetic cell modificationsin PGE2-treated (A) and HRG-treated (B) colorectal carcinoma cells.These early modifications correspond to a 6-h PGE2 treatment and a12-h HRG treatment. FITC-conjugated antimouse antibody stainingshows the distribution of uPA; Alexa546-conjugated antirabbit second-ary antibody indicates the localization ofb-catenin.C, confocal imagesshow late morphogenetic cell modifications (after 48 h of treatment)with lumens (arrow) inside the three-dimensional structures and block-age by interference with the uPA/uPAR pathway.D, Boyden chamberinvasion assay. Consistent with the effects of inhibitors in the three-dimensional assay, Å36 blocked the stimulation of FET cell invasion byboth PGE2 and HRG.

86




colon system. The results of our study suggest a model of colon cancerprogression in which the binding of HRG to its HER-3 and HER-4receptors on adjacent tumor epithelial cells induces a series of events,including secretion of PGE2 and up-regulation of uPA. This newpathway may be integrated into a broad model of how cells function,die, migrate, or differentiate under the influence of HRG and PGE2

signals. In addition, our three-dimensional model system presentedhere may potentially be used to screen the effects of therapeuticcompounds and biomolecules on different spatial aspects of colonbiology, including cell growth, motility, invasion, survival, and ap-optosis.

References

1. Prescott, S. M., and White, R. L. Self-promotion? Intimate connections between APCand prostaglandin H synthase-2. Cell,87: 783–786, 1996.

2. Smith, W. L., Garavito, R. M., and DeWitt D. L. Prostaglandin endoperoxide Hsynthase (cyclooxoygenase)-1 and -2. J. Biol. Chem.,271: 33157–33160, 1996.

3. Kutchera, W., Jones, D. A., Matsunami, N., Groden, J., McIntyre, T. M., Zimmerman,G. A., White, R. L., and Prescott, S. M. Prostaglandin H synthase 2 is expressedabnormally in human colon cancer: evidence for a transcriptional effect. Proc. Natl.Acad. Sci. USA,93: 4816–4820, 1996.

4. Kargman, S. L., O’Neill, G. P., Vickers, P. J., Evans, J. F., Mancini, J. A., and Jothy,S. Expression of prostaglandin G/H synthase-1 and -2 protein in human colon cancer.Cancer Res.,55: 2556–2559, 1995.

5. Eberhart, C. E., Coffey, R. J., Radhika, A., Giardielo, F. M., Ferrenbach, S., andDuBois, R. N. Up-regulation of cyclooxygenase 2 gene expression in human colo-rectal adenomas and adenocarcinomas. Gastroenterology,107: 1183–1188, 1994.

6. DuBois, R. N., Radhika, A., Reddy, B. S., and Entingh, A. J. Increased cyclooxy-genase-2 levels in carcinogen-induced rat colonic tumors. Gastroenterology,110:1259–1262, 1996.

7. Sheng, H., Shao, J., Kirkland, S. C., Isakson, P., Coffey, R. J., Morrow, J.,Beauchamp, R. D., and DuBois, R. N. Inhibition of human colon cancer cell growthby selective inhibition of cyclooxygenase-2. J. Clin. Investig.,99: 2254–2259, 1997.

8. Oshima, M., Dinchuk, J. E., Kargman, S. L., Oshima, H., Hancock, B., Kwong, E.,Trzaskos, J. M., Evans, J. F., and Taketo, M. M. Suppression of intestinal polyposisin APC716 knockout mice by inhibition of cyclooxygenase 2. Cell,87: 803–809,1996.

9. Tsujii, M., and DuBois, R. N. Alterations in cellular adhesion and apoptosis inepithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell,83:493–501, 1995.

10. Sheng, H., Shao, J., Morrow, J. D., Beauchamp, R. D., and DuBois, R. N. Modulationof apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells.Cancer Res.,58: 362–366, 1998.

11. Tsuji, M., Kawano, S., and Dubois, R. N. COX-2 expression in human colon cancercells increases metastatic potential. Proc. Natl. Acad. Sci. USA,94: 3336–3340,1997.

12. Lozano, Y., Taitz, A., Petruzzelli, G. J., Djordjevic, A., and Young, M. R. PGE2-protein kinase A signaling and protein phosphatases-1 and -2A regulate human head

and neck squamous cell carcinoma motility, adherence, and cytoskeleton organiza-tion. Prostaglandins,51: 35–48, 1996.

13. Hynes, N. C., and Stern, D. F. The biology of erbB-2/neu/HER-2 and its role incancer. Biochim. Biophys. Acta, 1198:165–184, 1994.

14. Adam, L., Vadlamudi, R., Kondapaka, S. B., Chemoff, J., Mendelsohn, J., andKumar, R. Heregulin regulates cytoskeletal reorganization and cell migration throughthe p21-activated kinase-1 via phosphatidylinositol-3 kinase. J. Biol. Chem.,273:28238–28246, 1998.

15. Vadlamudi, R., Mandal, M., Adam, L., Steinbach, G., Mendelsohn, J., and Kumar, R.Regulation of cyclooxygenase-2 pathway by HER2 receptor. Oncogene,18: 305–314,1999.

16. Mazar, A. P., Henkin, J., and Goldfarb, R. H. The urokinase plasminogen activatorsystem in cancer: implications for tumor angiogenesis and metastasis. Angiogenesis,3: 15–32, 1999.

17. Andreasen, P. A., Kjøller, L., Christensen, L., and Duffy, M. The urokinase-typeplasminogen activator system in cancer metastasis: a review. Int. J. Cancer,72: 1–22,1997.

18. Yamamoto, M., Sawaya, R., Mohanam, S., Rao, V. H., Bruner, J. M., Nicolson, G. L.,and Rao, J. S. Expression and localization of urokinase-type plasminogen activatorreceptor in human gliomas. Cancer Res.,54: 5016–5020, 1994.

19. Limongi, P., Resnati, M., Hernandez-Marrero, L., Cremona, O., Blasi, F., and Fazioli,F. Biosynthesis and apical localization of the urokinase receptor in polarized MDCKepithelial cells. FEBS Lett.,369: 207–211, 1995.

20. Pyke, C., Kristensen, P., Ralfkiaer, E., Grondahl-Hansen, J., Eriksen, J., Blasi, F., andDano, K. Urokinase-type plasminogen activator is expressed in stromal cells and itsreceptor in cancer cells at invasive foci in human colon adenocarcinomas. Am. J.Pathol.,138: 1059–1067, 1991.

21. Mizukami, I. F., Garni-Wagner, B. A., DeAngelo, L. M., Liebert, M., Flint, A.,Lawrence, D. A., Cohen, R. L., and Todd, R. F., III. Immunologic detection of thecellular receptor for urokinase plasminogen activator. Clin. Immunol. Immunopathol.,71: 96–104, 1994.

22. Buo, L., Meling, G. I., Karlsrud, T. S., Johansen, H. T., and Aasen, A. O. Antigenlevels of urokinase plasminogen activator and its receptor at the tumor-host interfaceof colorectal adenocarcinomas are related to tumor aggressiveness. Hum. Pathol.,26:1133–1138, 1995.

23. Mandal, M., Adam, L., Mendelsohn, J., and Kumar, R. Nuclear targeting of Baxduring apoptosis in human colorectal cancer cells. Oncogene,17: 999–1007, 1998.

24. McRea, P. D., Briecher, W. M., and Gumbiner, B. M. Induction of a secondary bodyaxis in Xenopusby antibodies tob-catenin. J. Cell. Biol.,123: 477–484, 1993.

25. Kumar, R., and Mendelsohn, J. Role of the 29-59-oligoadenylate synthetase inginterferon-mediated growth inhibition of A431 cells. Cancer Res.,49: 3482–3488,1989.

26. Mazumdar, A., Adam, L., Boyd, D., and Kumar, R. Heregulin regulation of plasmin-ogen activator and its receptor: human breast epithelial cell invasion. Cancer Res.,61:400–405, 2001.

27. Mazar, A. P., and Jones, T. R. Å 36 is RTLPPD. PCT Publication WO 00/26353, May11, 2000.

28. Fukamachi, H., and Kim, Y. S. Importance of desmosome formation for glandularorganization of LS174T human colon cancer cells in organ culture. Dev. GrowthDiffer., 31: 307–313, 1989.

29. Ilantzis, C., and Stanner, C. P. A colonic tissue architecture assay applied to humancolon carcinoma cells. In Vitro Cell. Dev. Biol. Animal,33: 50–61, 1997.

87




2001;61:81-87. Cancer Res Liana Adam, Abhijit Mazumdar, Tushar Sharma, et al. 2

EInvasiveness of Cancer Cells by Heregulin and Prostaglandin A Three-Dimensional and Temporo-Spatial Model to Study

Updated version

http://cancerres.aacrjournals.org/content/61/1/81

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/61/1/81.full#ref-list-1

This article cites 28 articles, 10 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/61/1/81.full#related-urls

This article has been cited by 6 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/61/1/81To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/content/61/1/81.full#ref-list-1

http://cancerres.aacrjournals.org/content/61/1/81.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



a three-dimensional and temporo-spatial model to study … · [cancer research 61, 81–87, january...

Documents