Mechanical model of cytoskeleton structuration during cell adhesion and spreading.

The biomechanical behavior of an adherent cell is intimately dependent on its cytoskeleton structure. Several models have been proposed to study this structure taking into account its existing internal forces. However, the structural and geometrical complexities of the cytoskeleton's filamentous networks lead to difficulties for determining a biologically realistic architecture. The objective of this paper is to present a mechanical model, combined with a numerical method, devoted to the form-finding of the cytoskeleton structure (shape and internal forces) when a cell adheres on a substrate. The cell is modeled as a granular medium, using rigid spheres (grains) corresponding to intracellular cross-linking proteins and distant mechanical interactions to reproduce the cytoskeleton filament internal forces. At the initial state (i.e., before adhesion), these interactions are tacit. The adhesion phenomenon is then simulated by considering microtubules growing from the centrosome towards transmembrane integrin-like receptors. The simulated cell shape changes in this process and results in a mechanically equilibrated structure with traction and compression forces, in interaction with the substrate reactions. This leads to a compressive microtubule network and a corresponding tensile actin-filament network. The results provide coherent shape and forces information for developing a mechanical model of the cytoskeleton structure, which can be exploitable in future biomechanical studies of adherent cells.

Ep-CAM transfection in thymic epithelial cell lines triggers the formation of dynamic actin-rich protrusions involved in the organization of epithelial cell layers.

Thymic epithelium is organized in a highly connected three-dimensional network through which thymocytes differentiate. The molecular mechanisms underlying this organization are still unknown. In thymic medulla, a major site of tolerance induction, the development of the epithelial cell net is tightly regulated by the needs of thymocyte selection. These reticulated epithelial cells express high levels of the Ep-CAM molecule. Using different thymic epithelial cell lines as a model system, we found that transfection of Ep-CAM enhances cell growth and leads to a rapid reorganization of the actin cytoskeleton by inducing the formation of numerous stress fibers and long cell protrusions. Finally, the crosslinking of the extracellular domain of a chimeric CD25ec/Ep-CAMic molecule is sufficient to trigger the formation of protrusions. These results suggest that expression of Ep-CAM might balance the organizing capacity of cadherin molecules and may be participating in the formation of a dynamic stromal cell network in the thymus.

Rabbit M cells and dome enterocytes are distinct cell lineages.

We have studied the M cell origin and differentiation pathway in rabbit gut-associated lymphoid tissues. Micro-dissected domes and epithelium isolated by ethylene diamine tetra acetic acid detachment allowed us to view the whole epithelial surface from the bottom of crypts to the top of domes. We used monoclonal antibodies specific to the apex of either M cells or dome enterocytes, lectins, and antibodies to vimentin in appendix, distal Peyer's patches and caecal patches. The earliest vimentin-labeled M cells were observed in the BrdU-positive proliferative zone of dome-associated crypts. Gradual differentiation of the M cell vimentin cytoskeleton started at this site to progressively give rise to the first pocket-forming M cells in the upper dome. Therefore, these mitotic cells of the crypts appear as the direct precursors of M cells. In addition to an early appearance of M cell markers, a regular mosaic-like relative distribution of M cells and dome enterocytes was already detected in the vicinity of crypts, similar to that observed on the lateral surface of domes where functional M cells lie. This constant distribution implies that there is no trans-differentiation of enterocytes to M cells along the crypt-dome axis. Together, these observations provide very strong evidence in favor of an early commitment in crypts of M cell and enterocyte distinct lineages.

Glycocalyx on rabbit intestinal M cells displays carbohydrate epitopes from Muc2.

It is essential to investigate the apical surface properties of both M cells and dome enterocytes to understand the mechanisms involved in the binding of pathogens to M cells. In rabbit appendix tissue, monoclonal antibodies (MAbs) highlight differences between M cells (MAb 58) and dome enterocytes (MAb 214). Such antibodies ultimately recognized intestinal mucin-related epitopes. To further characterize these differences, the labeling patterns obtained with these MAbs were compared to those obtained with other antibodies to intestinal mucins on dissected domes from all gut-associated lymphoid tissues. A glycoprotein recognized by MAb 58 was purified on a CsCl isopycnic density gradient and microsequenced, and its mRNA expression was localized by in situ hybridization. It was identified as the rabbit homologue of human Muc2, i.e., the major mucin secreted in intestine tissue. Two other Muc2 carbohydrate epitopes were also expressed on M cells, although Muc2 mRNA was not detected. All results indicated that M cells express, on their apical membrane, glycoconjugates bearing at least three glycosidic epitopes from Muc2. MAb 214 and MAb 6G2, which recognized a partially characterized mucin expressed on dome enterocytes, were negative markers for M cells in rabbit gut-associated lymphoid tissues. We propose that the presence, on the surface of M cells, of carbohydrates also expressed on Muc2, together with the absence of an enterocyte-associated mucin, could favor pathogen attachment and accessibility to the M-cell luminal membrane.

Mutagenesis of the phosphatidylinositol 4,5-bisphosphate (PIP(2)) binding site in the NH(2)-terminal domain of ezrin correlates with its altered cellular distribution.

The cytoskeleton-membrane linker protein ezrin has been shown to associate with phosphatidyl-inositol 4,5-bisphosphate (PIP(2))-containing liposomes via its NH(2)-terminal domain. Using internal deletions and COOH-terminal truncations, determinants of PIP(2) binding were located to amino acids 12-115 and 233-310. Both regions contain a KK(X)(n)K/RK motif conserved in the ezrin/radixin/moesin family. K/N mutations of residues 253 and 254 or 262 and 263 did not affect cosedimentation of ezrin 1-333 with PIP(2)-containing liposomes, but their combination almost completely abolished the capacity for interaction. Similarly, double mutation of Lys 63, 64 to Asn only partially reduced lipid interaction, but combined with the double mutation K253N, K254N, the interaction of PIP(2) with ezrin 1-333 was strongly inhibited. Similar data were obtained with full-length ezrin. When residues 253, 254, 262, and 263 were mutated in full-length ezrin, the in vitro interaction with the cytoplasmic tail of CD44 was not impaired but was no longer PIP(2) dependent. This construct was also expressed in COS1 and A431 cells. Unlike wild-type ezrin, it was not any more localized to dorsal actin-rich structures, but redistributed to the cytoplasm without strongly affecting the actin-rich structures. We have thus identified determinants of the PIP(2) binding site in ezrin whose mutagenesis correlates with an altered cellular localization.

Membrane recruitment of Rac1 triggers phagocytosis.

Rac1 is a &Rgr;-family GTP-binding protein that controls lamellipodia formation and membrane ruffling in fibroblasts. Recently, Rac1 and Cdc42, another member of the &Rgr;-family, have been shown to regulate Fc receptor-mediated phagocytosis in macrophages by controlling different steps of membrane and actin dynamics leading to particle engulfment. Here, we investigated the function of Rac1 using a membrane recruitment system that mimics phagocytosis. Recruitment of an activated Rac1 protein to the cytoplasmic domain of an engineered membrane receptor by using rapamycin as a bridge induces ingestion of latex beads bound to the receptor. Rac1-mediated bead uptake depends on actin polymerisation since actin filaments accumulate at the bead/membrane binding sites and internalisation is inhibited by cytochalasin D. Internalisation is also abolished upon substitution of Phe37 to Leu in the Rac1 effector region. Our results indicate that by promoting actin polymerisation at particle attachment sites, Rac1 by acting through specific downstream effectors induces plasma membrane remodeling that allows particle internalisation in a membrane-enclosed phagosome.

Inducible recruitment of Cdc42 or WASP to a cell-surface receptor triggers actin polymerization and filopodium formation.

BACKGROUND: Cdc42, a GTP-binding protein of the Rho family, controls actin cytoskeletal organization and helps to generate actin-based protruding structures, such as filopodia. In vitro, Cdc42 regulates actin polymerization by facilitating the creation of free barbed ends - the more rapidly growing ends of actin filaments - and subsequent elongation at these ends. The Wiskott- Aldrich syndrome protein, WASP, which has a pleckstrin-homology domain and a Cdc42/Rac-binding motif, has been implicated in cell signaling and cytoskeleton reorganization. We have investigated the consequences of local recruitment of activated Cdc42 or WASP to the plasma membrane. RESULTS: We used an activated Cdc42 protein that could be recruited to an engineered membrane receptor by adding rapamycin as a bridge, and added antibody-coupled beads to aggregate these receptors. Inducible recruitment of Cdc42 to clusters of receptors stimulated actin polymerization, resulting in the formation of membrane protrusions. Cdc42-induced protrusions were enriched in the vasodilator-stimulated phosphoprotein VASP and the focal-adhesion-associated proteins zyxin and ezrin. The Cdc42 effector WASP could also induce the formation of protrusions, albeit of different morphology. CONCLUSIONS: This is the first demonstration that the local recruitment of activated Cdc42 or its downstream effector, WASP, to a membrane receptor in whole cells is sufficient to trigger actin polymerization that results in the formation of membrane protrusions. Our data suggest that Cdc42-induced actin-based protrusions result from the local and serial recruitment of cytoskeletal proteins including zyxin, VASP, and ezrin.

Mucin-related epitopes distinguish M cells and enterocytes in rabbit appendix and Peyer's patches.

The biochemical composition of the apical membranes of epithelial M cells overlying the gut-associated lymphoid tissues (GALT) is still largely unknown. We have prepared monoclonal antibodies (MAbs) directed against carbonate-washed plasma membranes from epithelial cells detached with EDTA from rabbit appendix, a tissue particularly rich in GALT. As determined by immunofluorescence microscopy, several MAbs specifically recognized either M cells or enterocyte-like cells of the domes from rabbit appendix, sacculus rotundus, and Peyer's patches. M cells were identified by their large ventral pocket containing lymphoid cells and by specific labeling with antivimentin. Among various characterized MAbs, MAb 104 recognized rabbit immunoglobulins and was used as an apical marker for M cells in the rabbit appendix, MAb 58 selectively stained an integral membrane glycoprotein of greater than 205 kDa located at the apex of M cells, and MAb 214 stained a smaller soluble glycoprotein associated with the apical surfaces from neighboring enterocytes. In addition, both MAbs 58 and 214 also labeled luminal mucus and secretory granules in goblet cells. The selective association of mucin-related molecules at the surfaces of either M cells or enterocyte-like cells of the follicle-associated epithelium suggests that specific carbohydrate antigens are differentially expressed by epithelial cells and could account for the differential binding properties of pathogens.

Fc receptor-mediated phagocytosis requires CDC42 and Rac1.

At the surface of phagocytes, antibody-opsonized particles are recognized by surface receptors for the Fc portion of immunoglobulins (FcRs) that mediate their capture by an actin-driven process called phagocytosis which is poorly defined. We have analyzed the function of the Rho proteins Rac1 and CDC42 in the high affinity receptor for IgE (FcepsilonRI)-mediated phagocytosis using transfected rat basophil leukemia (RBL-2H3) mast cells expressing dominant inhibitory forms of CDC42 and Rac1. Binding of opsonized particles to untransfected RBL-2H3 cells led to the accumulation of F-actin at the site of contact with the particles and further, to particle internalization. This process was inhibited by Clostridium difficile toxin B, a general inhibitor of Rho GTP-binding proteins. Dominant inhibition of Rac1 or CDC42 function severely inhibited particle internalization but not F-actin accumulation. Inhibition of CDC42 function resulted in the appearance of pedestal-like structures with particles at their tips, while particles bound at the surface of the Rac1 mutant cell line were enclosed within thin membrane protrusions that did not fuse. These phenotypic differences indicate that Rac1 and CDC42 have distinct functions and may act cooperatively in the assembly of the phagocytic cup. Inhibition of phagocytosis in the mutant cell lines was accompanied by the persistence of tyrosine-phosphorylated proteins around bound particles. Phagocytic cup closure and particle internalization were also blocked when phosphotyrosine dephosphorylation was inhibited by treatment of RBL-2H3 cells with phenylarsine oxide, an inhibitor of protein phosphotyrosine phosphatases. Altogether, our data show that Rac1 and CDC42 are required to coordinate actin filament organization and membrane extension to form phagocytic cups and to allow particle internalization during FcR-mediated phagocytosis. Our data also suggest that Rac1 and CDC42 are involved in phosphotyrosine dephosphorylation required for particle internalization.

Selective control of membrane ruffling and actin plaque assembly by the Rho GTPases Rac1 and CDC42 in FcepsilonRI-activated rat basophilic leukemia (RBL-2H3) cells.

Engagement of the high affinity IgE receptor (FcepsilonRI) in mast cells elicits a series of intracellular signalling events including cytoskeletal reorganization and granule exocytosis. To analyze the coupling of receptor activation to specific cytoskeletal responses, we expressed dominant negative mutant forms of the Rho GTPases CDC42 and Rac1 in rat RBL-2H3 tumor mast cells. We show here that dominant inhibition of CDC42 function decreases cell adhesion, interferes with Fc(epsilon)RI-induced actin plaque assembly and reduced the recruitment of vinculin at the cell-substratum interface, while the inhibitory Rac1 mutant abolishes Fc(epsilon)RI-mediated membrane ruffling. The expression of trans-dominant inhibitory forms of either CDC42 or Rac1 significantly inhibited antigen-induced degranulation. Altogether, our results demonstrate that CDC42 and Rac1 control distinct pathways downstream of FcepsilonRI engagement leading either to the induction of actin plaques, or to the production of membrane ruffles. These two pathways are critically involved during the degranulation response induced by Fc(epsilon)RI aggregation.

Three determinants in ezrin are responsible for cell extension activity.

The ERM proteins--ezrin, radixin, and moesin--are key players in membrane-cytoskeleton interactions. In insect cells infected with recombinant baculoviruses, amino acids 1-115 of ezrin were shown to inhibit an actin- and tubulin-dependent cell-extension activity located in ezrin C-terminal domain (ezrin310-586), whereas full-length ezrin1-586 did not induce any morphological change. To refine the mapping of functional domains of ezrin, 30 additional constructs were overexpressed in Sf9 cells, and the resulting effect of each was qualitatively and semiquantitatively compared. The removal of amino acids 13-30 was sufficient to release a cell-extension phenotype. This effect was abrogated if the 21 distal-most C-terminal amino acids were subsequently deleted (ezrin31-565), confirming the existence of a head-to-tail regulation in the whole molecule. Surprisingly, the deletion in full-length ezrin of the same 21 amino acids provided strong cell-extension competence to ezrin1-565, and this property was recovered in N-terminal constructs as short as ezrin1-310. Within ezrin1-310, amino acid sequences 13-30 and 281-310 were important determinants and acted in cooperation to induce cytoskeleton mobilization. In addition, these same residues are part of a new actin-binding site characterized in vitro in ezrin N-terminal domain.

Breast cancer cells have a high capacity to acidify extracellular milieu by a dual mechanism.

The extracellular pH in malignant tumors is known to be lower than in normal tissues and may therefore facilitate extracellular activation of secreted lysosomal cathepsins. We have tested the capability of human mammary cells (continuous cell lines and primary culture) to acidify their extracellular environment, using two techniques. By measuring pH changes through alterations of phenolsulfone phthaleine absorbance, we found that the more aggressive MDA-MB-231 human breast cancer cells were more active in acidifying a non-buffered balanced salt solution than the estrogen receptor positive MCF7 and ZR75 cell lines and than normal mammary epithelial cells in primary culture. Metastatic breast cancer cells from pleural effusions were up to 200-fold more active in acidifying their extracellular milieu than non-malignant mammary cells cultured in the same conditions, strongly suggesting that this difference also occurs in vivo. The use of inhibitors in the presence or absence of glucose showed that both lactate and an ATP-driven proton pump sharing some characteristics of the vacuolar H+ pump were involved. Bafilomycin A1, a specific inhibitor of the vacuolar (V-type) ATP-H+ pump inhibited part of the acidification by MCF7 cells, but not by MDA-MB-231 cells. We also used microelectrodes to measure extracellular pH, in close contact to the MCF7 breast cancer cells. The pH at the free surface of MCF7 cells was lower by 0.33 +/- 0.14 unit than that of the surrounding medium, while insertion of the microelectrode tip beneath the attached surface of the cells showed a greater lowering of pH from 0.3 to 1.7 pH unit as long as cell attachment on the substrate prevented H+ diffusion. We conclude that breast carcinoma cells have a higher capacity for acidifying their extracellular milieu than normal mammary cells, and that both a plasma membrane H(+)-ATPase, and lactic acid production are involved in this acidification. It is therefore possible that the aspartyl and cysteinyl pro-cathepsins secreted in excess by tumor cells may be activated extracellularly in vivo close to the basement membrane.

Ezrin NH2-terminal domain inhibits the cell extension activity of the COOH-terminal domain.

Overexpression in insect cells of the full coding sequence of the human membrane cytoskeletal linker ezrin (1-586) was compared with that of a NH2-terminal domain (ezrin 1-233) and that of a COOH-terminal domain (ezrin 310-586). Ezrin (1-586), as well as ezrin (1-233) enhanced cell adhesion of infected Sf9 cells without inducing gross morphological changes in the cell structure. Ezrin (310-586) enhanced cell adhesion and elicited membrane spreading followed by microspike and lamellipodia extensions by mobilization of Sf9 cell actin. Moreover some microspikes elongated into thin processes, up to 200 microns in length, resembling neurite outgrowths by a mechanism requiring microtubule assembly. Kinetics of videomicroscopic and drug-interference studies demonstrated that mobilization of actin was required for tubulin assembly to proceed. A similar phenotype was observed in CHO cells when a comparable ezrin domain was transiently overexpressed. The shortest domain promoting cell extension was localized between residues 373-586. Removal of residues 566-586, involved in in vitro actin binding (Turunen, O., T. Wahlström, and A. Vaheri. 1994. J. Cell Biol. 126:1445-1453), suppressed the extension activity. Coexpression of ezrin (1-233) with ezrin (310-586) in the same insect cells blocked the constitutive activity of ezrin COOH-terminal domain. The inhibitory activity was mapped within ezrin 115 first NH2-terminal residues. We conclude that ezrin has properties to promote cell adhesion, and that ezrin NH2-terminal domain negatively regulates membrane spreading and elongation properties of ezrin COOH-terminal domain.

Specific mannose-6-phosphate receptor-independent sorting of pro-cathepsin D in breast cancer cells.

The secretion of pro-cathepsin D (pro-cath-D) in some human metastatic breast cancer cells (MCF7, MDA/MB231), contrary to normal mammary cells, is not increased by ammonium chloride treatment, indicating a mannose-6-phosphate-independent sorting to lysosomes. By studying a variety of cell lines and lysosomal enzymes, we show that secretion of newly synthesized pro-cath-D was not mediated by the 46-kDa mannose-6-phosphate receptor (MPR) and that its resistance to NH4Cl for secretion was specific to cath-D and not to other lysosomal enzymes. This resistance appeared to be correlated with the basal hypersecretion of pro-cath-D, but not with its overexpression. By contrast, pro-cath-D secretion was increased by NH4Cl in fibroblasts and nontumoral epithelial mammary cells, suggesting a specificity for cancer cells. Immunofluorescence staining showed that pro-cath-D, but neither cathepsin B nor beta-hexosaminidase, accumulated in intracytoplasmic vesicles of cells treated with ammonium chloride. In pulse--chase experiments and by subcellular fractionation on Percoll gradient, cath-D was found to be sorted into dense lysosomes whether cells were treated or not by NH4Cl. Treatment of cells with NH4Cl, however, inhibited processing and maturation of pro-cath-D, which was also observed in light vesicles in the absence of NH4Cl. Part of pro-cath-D, but not processed enzyme, was also found to be membrane associated in saponin-permeabilized cells. We conclude that in breast cancer cells, the MPR-independent pathway of pro-cath-D to lysosome is predominant compared to normal cells and other lysosomal enzymes. This alternative pathway should therefore be considered, in addition to MPR, to explain pro-cath-D sorting and activation in breast cancer cells.

Cathepsin D immunostaining in paraffin-embedded breast cancer cells and macrophages: correlation with cytosolic assay.

High cathepsin D (cath-D) concentration in breast cancer cytosol is associated with increased risk of metastasis. To specify the relative contribution of the different cells types responsible for cath-D level in cytosol, we validated semiquantitative cath-D immunoperoxidase staining on formalin-fixed, paraffin-embedded sections, using the M1G8 monoclonal antibody, one of the two antibodies of the cytosolic assay. Using computer-aided image analysis, cath-D level in cancer cells was estimated by integrating both staining intensity in each cell and proportion of stained cells. We confirmed on 41 primary breast cancers a higher expression of cath-D in cancer cells compared with peritumoral mammary glands. Cancer cell staining was mostly in lysosomes and for some invasive ductal carcinomas in large vesicles corresponding to phagosomes. Lymphocytes and fibroblasts were not or were only weakly stained. Macrophages also were stained for cath-D, generally on the periphery of the tumor area. The cytosolic cath-D level was correlated with cath-D expression in cancer cells (r = .76; P = 1 x 10(-4)) rather than with the number of macrophages in the tumor (r = .29; P = .09), as determined by use of the specific anti-CD68 antibody. There was a significant increase in the tissue cath-D level in tumors containing large vesicles compared with tumors without large vesicles. This approach provides a means to separately estimate the prognostic significance of cath-D expression in cancer cells and macrophages when evaluating risk of metastasis.

Characterization of very acidic phagosomes in breast cancer cells and their association with invasion.

Human metastatic breast cancer cells in culture contain large acidic vesicles (diameter 5-10 microns) in which endocytosed extracellular matrix can be digested by activated lysosomal proteinases such as cathepsin D (P. Montcourrier et al. (1990). Cancer Res. 50, 6045-6054). We examined these large compartments by transmission electron microscopy, measured their pH by video-enhanced epifluorescence using FITC-dextran, and studied their functional significance. Their presence in metastatic MDA-MB231 cells was found to be correlated with an increased ability of cells to migrate through Matrigel and a high cathepsin D concentration. These cells were able to phagocytose 1.24 microns diameter latex beads and fluorescence Matrigel and incorporate this extracellular material into large acidic vesicles. This indicated that large acidic vesicles were associated with both phagocytosis and invasion, and are heterophagolysosomes rather than autophagosomes. Large acidic vesicles were actively acidified with a H(+)-ATPase vacuolar pump specifically inhibited by bafilomycin A1, and reached pH values (< 4), lower than the lysosomal value (pH approximately 5) in the same cells and in specialized phagocytotic cells such as macrophages. We conclude that the phagocytotic activity of breast cancer cells, associated with high cathepsin D expression, and high acidification potential, characterize cancer cells that have migrated through Matrigel.

[The presence of large acidic vesicles is correlated with in vitro invasive potential of breast cancer cells]. / La présence de grandes vésicules acides est corrélée avec le pouvoir invasif in vitro de cellules mammaires cancéreuses.

Human breast cancer cells include large acidic vesicles (LAVs) containing endocytosed extracellular matrix (P. Montcourrier et al., Cancer Res., 50, 1990, p. 6045-6054) and mature cathepsin D, a lysosomal protease associated with metastatic risk. We show that metastatic breast cancer MDA-MB231 cells migrating in vitro through a Matrigel gel in a modified Boyden chamber were enriched in LAVs (28% compared to 6% before plating) as visualized by acridine orange fluorescence. This is the first evidence that LAVs are associated with invasiveness of cancer rather than with an apoptotic process.

Characterization of two monoclonal antibodies against the COOH-terminal part of the human epidermal growth factor receptor and potential clinical use.

Two monoclonal antibodies of the IgG2 subclass, designated A 01 and B 11, were prepared against two synthetic peptides corresponding to the COOH-terminal sequence of the human epidermal growth factor receptor (EGF-R), in order to detect EGF-R in a radioimmunometric assay and by immunohistochemistry. Characterization of these Mabs showed that they recognized two different eptiopes on the original peptides with Kd of 1.7 x 10(-8) M and 1.3 x 10(-7) M, respectively, without crossreaction. The A 431 antigen recognized by A 01 and B 11 had an apparent molecular weight of approximately 170,000 and was able to specifically link to EGF. Thus, A 01 and B 11 are directed against an antigenic site on the human EGF-R. With Western blot analysis and immunostaining, A 01 was shown to be EGF-R specific. In addition to the EGF-R, B 11 recognized two unidentified soluble proteins present in the cytoplasm of the SKBR-3 cell line but different from the c-erb B-2 oncoprotein expressed by these cells. Mabs A 01 and B 11 were used in an IRMA for the determination of EGF-R using the A 431 cell line as a source of EGF-R. Mab A 01 was also shown to be a useful tool for immunohistochemical detection of EGF-R.

MCF7 mammary cancer cells respond to bFGF and internalize it following its release from extracellular matrix: a permissive role of cathepsin D.

High and low affinity receptors for basic fibroblast growth factor (bFGF) were detected by binding experiments on MCF7 breast cancer cells. These cells were stimulated for growth by physiological concentrations of bFGF. However, in contrast to endothelial cells, these MCF7 cells did not produce detectable amounts of biologically active bFGF or related heparin-binding growth factor(s) of the FGF family. In vitro, the cathepsin D (cath-D) secreted by MCF7 cells was able to digest extracellular matrix (ECM) and to release ECM-bound 125I-bFGF. When MCF7 cells were cultured on ECM containing bound bFGF, they internalized bFGF, which was slowly and partially proteolyzed in the cells. Processing occurred in acidic compartments and was inhibited by leupeptin. Pepstatin A, an inhibitor of aspartyl proteases, had no effect on the processing but reduced internalization of matrix-bound bFGF by MCF7 cells. Taken together, these results suggest a cooperation between cath-D and bFGF, by which the protease could facilitate the release of bFGF from ECM and its subsequent use by breast cancer cells and/or adjacent cells involved in angiogenesis.

Cathepsin D in breast cancer cells can digest extracellular matrix in large acidic vesicles.

In breast cancer cell lines, pro-cathepsin D is synthesized in excess and abnormally processed, resulting in its slower maturation and increased secretion into the culture medium. Since this lysosomal protease is only active at acidic pH, we have searched for acidic compartments other than lysosomes where cathepsin D might be active when MCF7 cells are plated on corneal extracellular matrix. We found large acidic intracellular vesicles (1.5 to 20 microns in diameter) by acridine orange and 3-(2,4-dinitroanilino)-3'-amino-N-methyldipropylamine staining, two fluorescent probes which reveal acidic compartments. These vesicles were actively acidified. They were 2- to 20-fold more abundant in MCF7 breast cancer cells and primary cultures of human breast cancers cells than in primary cultures of normal mammary epithelial cells. In living MCF7 cells, high resolution video-enhanced microscopy showed that these vesicles were mobile and intracellular. Double immunolocalization indicated that they contained mature cathepsin D (but no detectable pro-cathepsin D) and endocytosed extracellular material. This material (dextran, transferrin, and extracellular matrix) and the association with other lysosomal enzymes varied according to the vesicles, suggesting their heterogeneity (large endosomes or phagosomes). We conclude that, in breast cancer cells, cathepsin D may digest intracellularly phagocytosed and/or endocytosed extracellular matrix in large acidic vesicles. We propose that the higher expression of cathepsin D associated with the increased number of large acidic vesicles in breast cancer cells may facilitate digestion of basement membrane and consequently metastasis.