Skip to main content
NCBI home page
The American Journal of Pathology logoLink to The American Journal of Pathology
. 2004 Jun;164(6):1925–1933. doi: 10.1016/S0002-9440(10)63753-4

Inhibition of Proprotein Convertases Enhances Cell Migration and Metastases Development of Human Colon Carcinoma Cells in a Rat Model

Mimoun Nejjari *, Virginie Berthet *, Véronique Rigot *, Sullivan Laforest *, Marie-France Jacquier †, Nabil G Seidah ‡, Lionel Remy †, Erik Bruyneel §, Jean-Yves Scoazec †, Jacques Marvaldi *, José Luis *
PMCID: PMC1615749  PMID: 15161629

Abstract

Although proprotein convertases are involved in tumor development, nothing is known about their role in metastatic dissemination. To investigate the involvement of convertase inhibition, we used human colon carcinoma cells overexpressing α1-antitrypsin Portland (α1-PDX, PDX39P cells), a potent convertase inhibitor. We previously reported that these cells bear uncleaved integrin α subunits and display an altered attachment to vitronectin that is correlated with defects in the intracellular signaling pathways activated by αvβ5 integrin ligation. In this study, we demonstrate that the inhibition of proprotein convertase activity either by overexpression of α1-PDX or with the synthetic inhibitor decanoyl-Arg-Val-Lys-Arg-chloromethylketone (dec-RVKR-cmk) led to a significant increase in cell migration supported by the αvβ5 integrin. A collagen gel invasion assay showed that PDX39P cells also displayed an invasive ability, contrary to control cells. Moreover, when injected to immunosuppressed newborn rats, PDX39P cells were highly invasive, as they induce 10 times more metastases than mock-transfected cells. In addition, the aggressiveness of PDX39P cells can be greatly reduced by a function-blocking monoclonal antibody (mAb) against the αv subunit. It thus seems that inhibition of proprotein convertases enhances the in vivo invasiveness of colon tumor cells likely due to an increase in cell migration mediated by αv integrins.

The acquisition of cell motility and the capacity to invade basement matrix membranes and adjacent tissues play a central role in the complex multi-step process of metastasis. Cell migration is the result of dynamic interactions between the cell, the extracellular matrix (ECM), and the cytoskeleton. Integrins connect the ECM proteins outside to the actin cytoskeleton within the cell, allowing the traction required for cell migration.1,2 Integrins also play a crucial role in signal transduction events that control anchorage-independent growth and survival of tumor cells (for reviews, see references3–5). Therefore, these adhesion molecules play a pivotal role in tumor cell invasion and metastasis.5–7

Integrins are heterodimeric proteins composed by the non-covalent association of α and β subunits. Some integrin α chains undergo a post-translational cleavage in their extracellular domain by furin and PC5A, two proprotein convertases (PCs) of the subtilisin/kexin family.8,9 The role of this cleavage in integrin function remains unclear. Arguably, this post-translational processing of the α chain is not required for ligand binding,10–12 but is essential for the signaling function of α6β1 and αvβ5 integrins.10,13 Moreover, recent data suggest that the α6 subunit cleavage might be linked to the differentiation state of lens cells.14

PCs are responsible for the biological activation of a variety of precursor proteins, including pro-hormones and their receptors and adhesion molecules. They are ubiquitously expressed and are involved in many physiological and pathological processes.15 The potential clinical and pharmacological role of the convertases fostered the development of several inhibitors. The most promising protein-based specific inhibitor of PCs is an α1-antitrypsin variant known as α1-antitrypsin Portland (α1-PDX). α1-PDX is a selective inhibitor of furin and, to a lesser extent, of PC5B16 and blocks the convertase-dependent processing of various precursors, including integrins.10,17–19

Recent reports have suggested a possible role for PCs in tumorigenesis (for reviews, see references15,20). Thus, furin expression levels correlate with invasiveness in human head and neck tumors and cell lines.21 Moreover, further data show that tumors obtained after subcutaneous inoculation of furin-overexpressing cells are larger and develop earlier than the controls.22 Conversely, inhibition of PCs resulted in reduced in vivo tumorigenicity and in vitro cell proliferation.23–25

We have recently reported that the absence of endoproteolytic cleavage of αv integrins, by overexpression of α1-PDX in colonic cancer cells, has important consequences on signal transduction pathways leading to alterations in integrin function such as cell adhesion.10 PCs inhibition also leads to reduced cell proliferation in vitro and delayed tumor growth in nude mice inoculated with tumor cells.23 However, herein we demonstrate that PCs inhibition moderately reduced subcutaneous tumor growth but dramatically enhanced metastasis in immunosuppressed newborn rats. This aggressive behavior is likely due to an important increase in αv integrin-dependent cell migration.

Materials and Methods

Reagents

Dulbecco’s modified Eagle’s medium (DMEM) was purchased from Gibco (Cergy-Pontoise, France) and fetal calf serum (FCS) from BioWhittaker (Fontenay-sous-Bois, France). The furin inhibitor decanoyl-Arg-Val-Lys-Arg-chloromethylketone (dec-RVKR-cmk) was from Bachem (Voisins-le-Bretonneux, France). Protease inhibitors were from Sigma (St. Louis, MO). 35S-methionine was provided by Amersham (Les Ulis, France). Laminin-1 and vitronectin were prepared according to Timpl et al26 and Yatogho et al,27 respectively.

Rat mAb 69.6.5 against human αv integrin subunit was produced in our laboratory.28 Mouse monoclonal antibodies (mAbs) VNR139 (anti-αv subunit) and P1F6 (anti-αvβ5) and polyclonal antibody anti-HGF receptor were from Chemicon (Temecula, CA). Mouse C3VLA3 antibody against α3 integrin subunit was purchased from Immunotech (Marseille, France). Horseradish peroxidase (HRP)-conjugated secondary antibodies were from Amersham.

Cell Migration

Human IGROV1 and HT29-D4 cells, derived from human ovarian and colonic adenocarcinomas, respectively, were cultured as described.10,29 Control (empty vector, PDX0) and stably α1-PDX-transfected (PDX39P) HT29-D4 cells were described as previously reported.10

In vitro haptotaxis assays were performed using modified Boyden chambers (Transwell, Costar, Cambridge, MA), as previously described.30 In some experiments, cells were cultured for 48 hours in the presence of increasing concentrations of the furin inhibitor dec-RVKR-cmk. Cells were then metabolically labeled with 1 μCi/ml 35S-methionine for 2 hours and tested for migration in the presence of phorbol 12-myristate 13-acetate (PMA). Following incubation at 37°C, cells on the upper surface of the membrane were wiped with a cotton swab. Radioactivity of the migrated cells in the lower surface was determined in a liquid scintillation counter.

Collagen Invasion Assay

Collagen invasion assays were performed as previously described.31 Type I collagen was dissolved in bicarbonate buffer-containing DMEM neutralized with 1 mol/L NaOH. Samples (1.2 ml) were poured into a 6-well plate and incubated overnight at 37°C for gelation. Cells were harvested using Moscona and trypsin/EDTA buffers, and seeded on top of the collagen gels. After incubation for 24 hours at 37°C in the presence or absence of 10 nmol/L PMA, the depth of cell migration inside the collagen was measured using an inverted microscope. Invasive and superficial cells were counted in 12 predetermined fields of 0.157 mm2. The invasion index is the percentage of cells invading the gel over the total number of cells. DHD-FIB and MCF7/AZ cells were used as positive and negative controls of cell invasion, respectively.

In Vivo Tumor Induction

The Wistar newborn rats (six per cell line) received a ventral subcutaneous injection of 106 cells suspended in 100 μl phosphate-buffered saline (PBS). All rats were subsequently immunosuppressed by dorsal subcutaneous injections of anti-thymocyte serum32 at days 0, 2, 7, 14 and maintained in a specific pathogen-free environment throughout the experiment. Three weeks later, animals were sacrificed and the tumors and lungs were excised and processed for further analyses. The number of metastases of each lung was macroscopically and microscopically determined by three experimenters.

Reverse-Transcription/Polymerase Chain Reaction (RT-PCR)

Total RNA was extracted from tumors, metastases, and cultured cells using Trizol (Gibco). cDNAs were amplified by PCR, as already described,23 using the following oligonucleotides:

oligonucleotides (sense) oligonucleotides (anti-sense)

α1-PDX TCTTCTTCCTGCCTGATGAGG TAATACGACTCACTATAGGG

GAPDH TGGAAATCCCATCACCATCT GTCTTCTGG-GTGGCAGTGAT

The transcript of the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was taken as reference.

Immunoprecipitation and Immunoblot

PDX0 and PDX39P tumors were cut in small pieces and ground with an UltraTurax in 20 mmol/L Tris-HCl (pH 8), 200 mmol/L NaCl, 1 mmol/L EDTA, and 1% Triton X100 (RIPA buffer) containing a mixture of protease inhibitors (1 mmol/L PMSF, 500 U/ml aprotinin, 1 μg/ml leupeptin, 1 μmol/L pepstatin, 1 mmol/L iodoacetamide, and 1 mmol/L o-phenanthrolin). Cells were directly solubilized with the buffer. In some experiments, lysates were clarified by centrifugation and incubated with 5 μg rat mAb 69.6.5 overnight at 4°C and then with protein G-agarose for 45 minutes. Pellets were washed three times with RIPA buffer, three times with RIPA buffer/500 mmol/L NaCl, and once with PBS. Cell lysates or immunoprecipitated proteins were resolved by SDS-PAGE and blotted onto a nitrocellulose sheet. Membranes were blocked in PBS/5% nonfat milk/0.2% Tween 20 and probed overnight at 4°C with anti-αv mAb VNR139 (1:1000) or anti-HGF receptor (1:500). Blots were then revealed by chemiluminescence using HRP-conjugated secondary antibodies.

Results

Proprotein Convertases Inhibition Increases Cell Migration

We previously reported that cells expressing high levels of the PCs inhibitor α1-PDX (PDX39P cells) showed a reduced attachment to vitronectin.10 To determine whether α1-PDX expression also affects cell migration we used haptotaxis assays toward attractive proteins in modified Boyden chambers. As matrix proteins, we used vitronectin, the sole ligand for the integrin αvβ5 whose cleavage appears to be important for its function, and laminin-1 as a cleavage-independent substrate.10 As already reported for the parental cell line,30 both ECM proteins allow protein kinase C (PKC)-dependent migration of transfected cells (Figure 1). On stimulation with the phorbol ester PMA, the ability of PDX39P cell to migrate toward vitronectin was enhanced by more than threefold when compared to mock-transfected PDX0 cells (Figure 1A) or to the parental HT29-D4 cells (not shown). Moreover, ligand concentration allowing maximal migration was slightly higher for PDX39P (8 μg/ml) than for PDX0 (2 μg/ml). On the contrary, laminin-1 supported cell migration with the same efficiency for all cell types.

Figure 1.

Figure 1

Open in a new tab

Effect of α1-PDX expression on PMA-induced cell migration. A: Cell motility toward vitronectin or laminin-1 was determined using modified Boyden chambers. The bottom surface of porous membrane was pre-coated with the indicated concentrations of purified ECM protein. PDX0 (open diamonds) and PDX39P (closed circles) cells were seeded into the upper reservoir with 10 nmol/L PMA and allowed to migrate toward the lower reservoir for 5 hours at 37°C. Cells that migrated to the underside of the chamber were stained with Coomassie blue and enumerated. B: Cell migration toward 2.5 μg/ml vitronectin was performed as described above in the absence or presence of 10 μg/ml of function-blocking anti-αvβ5 (P1F6) or anti-α3 antibodies. Data are means ± SD of two independent experiments performed in triplicate.

Integrins are the major adhesion receptors, but integral membrane proteoglycans are also present at the cell surface and can bind ECM components. However, in our case, migration to vitronectin was dependent on αvβ5 integrins. Indeed, as shown on Figure 1B, the function-blocking mAb P1F6, against the αvβ5 integrin, prevented cell motility of both PDX0 and PDX39P cells toward vitronectin, while a mAb blocking the α3 subunit failed to affect cell migration.

To ascertain that the difference in behavior was not due to the selection process of transfected cells, PDX0 cells were treated for 48 hours with increasing concentrations of the furin inhibitor dec-RVKR-cmk before performing the migration assay. As illustrated in Figure 2A, migration of PDX0 cells to vitronectin increased with inhibitor concentration. The same result was obtained with laminin-5 which is another ECM protein recognized by cleavable α integrin subunits in PDX cells (unpublished results). Immunoblot confirmed that dec-RVKR-cmk indeed impaired, almost completely, the cleavage of αv subunit on PDX0 cells (Figure 2B). The αv-containing integrins have been involved in ovarian tumor progression. We therefore repeated the experiment with the ovarian cell line IGROV1 whose migration to vitronectin is αv-dependent.29 As shown in Figure 2C, incubation of IGROV1 cells with dec-RVKR-cmk also increased cell migration to vitronectin.

Figure 2.

Figure 2

Open in a new tab

PCs inhibition leads to increased cell migration. A: PDX0 cells were treated with various concentrations of the furin inhibitor dec-RVKR-cmk for 48 hours at 37°C. After 35S-methionine labeling, motility of PDX39P and PDX0 cells toward vitronectin was measured in modified Boyden chambers in the presence of 10 nmol/L PMA. B: Proteins from PDX0 cells treated (+) or not (−) for 48 hours with 100 μmol/L dec-RVKR-cmk were solubilized and resolved by SDS-polyacrylamide gel electrophoresis under reducing conditions. After blotting, the nitrocellulose membrane was probed with the anti-αv antibody VNR139. Uncleaved (αvNC) and cleaved (αvC) αv integrin subunits are emphasized. C: IGROV1 cells were treated with various concentrations of dec-RVKR-cmk for 48 hours at 37°C. Migration assays were performed, as described in Figure 1, using 8 μg/ml vitronectin. Results (± SD), expressed as the percentage of untreated cells (control), are from a representative experiment of three performed in duplicate.

Taken together, the results described above clearly demonstrate that inhibition of PCs leads to an enhanced cell migration. Migration of cells used in this study requires activation of PKC, either by PMA or by growth factors.30,33 Hepatocyte growth factor (HGF) is produced by various cell types, including tumor-bordering myofibroblasts, and functions locally as a cell migration inductor. As illustrated in Figure 3A, HGF promoted cell migration to vitronectin more efficiently than PMA. It is worth noting that, as observed on PMA stimulation, PDX39P cells are more motile than PDX0 in response to a physiological inductor such as HGF. This is not due to a higher expression of the HGF receptor, as both cells expressed similar amounts of receptor (Figure 3B). Interestingly, this receptor, which is a PCs substrate,34 was cleaved in PDX39P cells, indicating that other PCs than PC5A or furin may be involved in its endoproteolytic processing. This also demonstrates that the enhanced effect of HGF on cell migration is not due to a defect in receptor maturation.

Figure 3.

Figure 3

Open in a new tab

HGF and uPA induce cell migration. A: Migration of 35S-methionine-labeled PDX0 (filled bars) and PDX39P (open bars) cells was measured in modified Boyden chambers in the presence of PMA, HGF, or uPA. Results (± SD) are from a representative experiment of two performed in triplicate. B: Proteins from PDX cells were resolved by SDS-polyacrylamide gel electrophoresis under reducing (R) or non-reducing (NR) conditions. After blotting, the nitrocellulose membrane was probed with an antibody against the β chain of the human HGF receptor.

It has also been reported that PKC controls cell motility by regulating the expression of the plasminogen activator uPA. When exposed to uPA, PDX39P cells migrate more readily than control cells toward vitronectin (Figure 3A). It thus appears that whatever the stimulus used, the α1-PDX-expressing cells displayed an increase in cell migration compared to control cells.

Expression of α1-PDX Increases the Invasive Potential

The invasive potential of PDX cells was assessed in vitro in gels of type I collagen, the main component of ECM. Results were quantified and expressed as percentage of invasive cells in the collagen layer. As shown in Figure 4, stimulation of HT29-D4 and PDX0 cells by PMA did not alter their invasive potential. On the contrary, invasion of collagen gels by PDX39P cells was dramatically enhanced on PMA treatment. These results show that expression of the PCs inhibitor α1-PDX leads to a marked increase in in vitro invasion when cells are stimulated by PMA, as observed above for cell migration.

Figure 4.

Figure 4

Open in a new tab

Effect of α1-PDX expression on collagen gel invasion. Cells were allowed to invade type I collagen gels in the absence (open bars) or the presence (filled bars) of PMA for 24 hours at 37°C. The invasion index is the percentage of cells invading the gel. Results (± SD) are from a representative experiment of three performed in triplicate.

Proprotein Convertases Inhibition Increases Metastases Formation

To determine whether increased mobility and invasiveness of PDX39P cells could lead to an enhanced in vivo aggressiveness, we analyzed the ability of cells to develop metastases in an experimental animal model widely used for digestive cancers.35,36 PDX0 and PDX39P cells were inoculated into two groups of immunosuppressed newborn rats, and tumors and lungs were excised for further analyses after 21 days. Although similar in appearance (Figure 5A), PDX39P-induced tumors were 30% smaller compared to PDX0-induced tumors (Figure 5B). As seen in Figure 5C, rats inoculated with PDX0 cells developed a few lung metastases. Conversely, lungs from PDX39P cell-inoculated rats were colonized by a very high number of large metastases, leading to a dramatic increase in lung volume. This was confirmed by metastases quantification (Figure 5D). The aggressiveness of PDX39P cells led to the death of 17% of animals inoculated with these cells before the experiment completion. The autopsy of dead animals revealed a massive colonization of lung tissues by tumor cells (Figure 5E). This is in accordance with the fact that micrographs of PDX39P-induced tumors showed morphological evidence of higher local invasiveness and infiltrative pattern (invasion of adjacent tissues, vascular invasion, and tumor cell emboli) when compared to PDX0-induced tumors (data not shown).

Figure 5.

Figure 5

Open in a new tab

α1-PDX expression increases metastases formation. A: Macroscopic appearance of PDX39P- (a and b) and PDX0-induced tumors (c and d) 3 weeks after subcutaneous inoculation of cells in immunosuppressed newborn rats. B: Tumor volume was evaluated according to Stragand method. Data are mean values ± SD from two experiments (total of 12 rats per cell line). C: Macroscopic appearance of the lungs from PDX39P- (a and b) and PDX0-inoculated rats (c and d) 3 weeks after injection of tumor cells. D: Quantification of the number of lung metastases in PDX39P- and control cells-inoculated animals was done on each lung by three experimenters. Data represent mean values ± SD from two experiments (total of 12 rats per cell line). E: Masson trichrome staining of paraffin-embedded section. Micrograph (magnification, ×200) shows voluminous metastases (M) with little healthy tissue. Pneumocytes (arrows) are pushed back beneath the pressure of metastatic cells.

To ascertain that PDX39P cells still expressed α1-PDX after inoculation into rats we used two approaches. First, expression of the α1-PDX gene was assessed by semi-quantitative RT-PCR analysis in inoculated cells, as well as in several randomly chosen tumors and their corresponding metastases. The antisense oligomer was chosen from the vector promoter sequence to distinguish from endogenous α1-antitrysin. All but one of the PDX39P tumors tested were positive, indicating that expression of α1-PDX persisted during tumor growth (Figure 6A). More interestingly, all metastases developed by rats inoculated with PDX39P cells were α1-PDX positive. Second, we checked for the cleavage status of αv integrins in cells, tumors, and metastases by Western blot using a mAb specific for the human subunit. As illustrated in Figure 6B, only cell lysates from PDX39P tumors and their corresponding metastases displayed uncleaved αv subunits, suggesting that the expression level of α1-PDX was sufficient to impede, at least partially, integrin processing.

Figure 6.

Figure 6

Open in a new tab

α1-PDX expression inPDX39P-induced tumors and metastases. A: Expression of the α1-PDX gene was assessed by semi-quantitative RT-PCR analysis of total RNA extracted from inoculated cells (C), and from randomly chosen tumors (T) and their corresponding metastases (M). A gene product with the expected size is emphasized by an arrow. Bp, bp. B: The αv integrin subunit, extracted from cells (C), tumors (T) or metastases (M), was immunoprecipitated with mAb 69.6.5 and resolved by SDS-polyacrylamide gel electrophoresis under reducing conditions. After Western blot, the nitrocellulose membrane was probed with the anti-αv antibody VNR139. Uncleaved (αvNC) and cleaved (αvC) αv integrin subunits are emphasized. Analysis was performed on five (PDX0) or six (PDX39P) tumors and at least one of their corresponding metastasis. For the sake of clarity, only one sample of each category is shown.

Aggressiveness of PDX39P Cells Is αv-Dependent

The malignancy of tumor cells is essentially due to their invasive and metastatic potential, which needs integrin-dependent cell migration. To check the role of αv integrins in tumor cell aggressiveness, inoculation of cells into rats was performed in association with a function-blocking anti-αv mAb (69.6.5). The presence of the anti-integrin antibody led to a strong reduction in the number of lung metastases in PDX39P-inoculated rats (Figure 7; compare with Figure 5). Note that the mAb used in this experiment has been raised in rat and only blocked human αv integrins from tumor cells and not rat integrins from stromal or vascular cells. This experiment shows that the activity of αv integrins present on the tumor cell surface is essential for the invasiveness of colon carcinoma cells.

Figure 7.

Figure 7

Open in a new tab

Effect of anti-αv integrin mAb on metastases formation. A: PDX0 and PDX39P cells were injected subcutaneously into immunosuppressed newborn rats (six per cell line) along with 50 μg of the function-blocking rat anti-αv mAb 69.6.5. The antibody was then injected along the anti-thymocyte serum at days 2, 7, and 14. Photographs represent lungs excised from animals on day 21 after injection of cells. B: The number of lung metastases was macroscopically and microscopically quantified by three independent experimenters. Values represent means ± SD.

Discussion

PCs represent interesting potential targets for the development of new therapeutic agents. However, the efficient development and clinical application of PCs inhibitors require a better understanding of their molecular mechanisms in vitro and their impact on tumor growth and metastasis in various animal models before testing their efficacy in human disease. We, and others, previously observed that PCs inhibition leads to delayed tumor development and reduced tumor volume in nude or immunodeficient mice.22–25 However, herein we used immunosuppressed newborn rats to test the metastatic dissemination. Results obtained with this animal model show that the role of PCs in tumor development and progression is far from being as simple as suggested by previous observations.

As previously reported for nude mouse,23 the expression of the convertase inhibitor α1-PDX slightly decreased the tumorigenicity of human colonic PDX39P cells in immunosuppressed newborn rats. The reduced size of subcutaneous tumors in response to α1-PDX expression is likely due to differences in cell proliferation, as already reported in in vitro studies.22–24 By histological staining with an antibody against Ki67, a nuclear protein expressed in proliferating cells, we confirmed that in situ cell proliferation was higher in PDX0- than in PDX39P-induced tumors (data not shown).

However, despite their lower growth rate and the smaller tumors they produced, α1-PDX-expressing cells exhibit a very aggressive behavior when injected to immunosuppressed rats. Indeed, PDX39P-induced tumors showed morphological evidence of higher local invasiveness and infiltrative pattern and produced about 10 times more metastases than PDX0-induced tumors.

Among other processes, metastatic spread requires ECM remodeling, a reduction in cell adhesion, and an increase in cell motility, conditions fulfilled by PDX39P cells. The invasive behavior of PDX39P cells observed in vivo is associated with an increased cell migration toward vitronectin and with the invasion of collagen gels in vitro, which may be explained by alterations in cytoskeleton remodeling reported in a prior study.37 The motile and invasive behavior likely involves integrin αvβ5, the sole receptor for vitronectin in HT29-D428 because a function-blocking mAb against the αv subunit was efficient to prevent both in vivo invasion and in vitro cell motility (see Figures 1B and 7). Besides, the enhanced motility is clearly due to PCs inhibition (and not to the selection of one or a few clones of PDX39P during passaging) because incubation of PDX0 cells with a synthetic furin inhibitor, which prevents integrin cleavage, led to increased migration. Furthermore, PCs inhibition also enhanced motility of the ovarian cell line IGROV1 whose migration to vitronectin is αv-dependent.29 It also should be emphasized that the increased motile behavior of α1-PDX-expressing cells was observed no matter which migration inductor we tested (PMA, HGF, or uPA).

All α integrin subunits expressed at the cell surface in PDX39P cells lack endoproteolytic cleavage. This leads to a poor attachment of PDX39P cells to substrates recognized exclusively by cleavable α subunits and to a defect in the FAK signaling pathway.10 It is notable that, despite the reduced adhesion they allow, these same ECM proteins (ie, vitronectin and laminin-5) support the increased motility of PDX39P cells. It is worth noting that the inhibitor expression persisted on inoculation of PDX39P cells into immunosuppressed rats, as α1-PDX mRNA was present in PDX39P-induced tumors and metastases. Moreover, the expression level of the inhibitor was high enough to partially impede αv subunit cleavage. This suggests that the absence of endoproteolytic cleavage of the αv integrin subunit may be responsible for the observed increase in cell migration and aggressiveness. In support of this hypothesis, we observed that the blockage of αv subunit processing, by site-directed mutagenesis of the cleavage site, led to an invasive behavior on collagen on transfection in αv-deficient melanoma cells M21L. This work will be described in detail elsewhere (V. Berthet, manuscript in preparation).

As can be seen in Figure 1 and already reported,30 cell migration could only be observed in a narrow range of vitronectin concentration. Such biphasic curves in cell migration have also be reported for other cell lines.29,38 As cell migration requires the formation of new attachments at the cell front and the break of attachments at the rear, maximum cell speed is predicted to occur at an intermediate ratio of cell-substratum adhesiveness to intracellular contractile force.39 Low ligand concentrations (or integrin activity) decrease the probability of formation of stable attachments by lamellipodes, while high ligand concentrations (or integrin activity) probably hinder cell migration by obstructing the release of adhesions at the rear. The reduced adhesion due to the absence of proteolytic processing of αv subunit10 could thus, at least partly, explain the increased motility of PDX39P cells.

The molecular mechanism by which the endoproteolytic cleavage of αv subunit affects cell migration and aggressiveness is not precisely understood. Potential hypotheses include alterations in the cellular proteins associated with or regulated by αv integrins. Thus, the unconventional processing of αv by membrane type 1 matrix metalloproteinase (MT1-MMP), and the subsequent generation of the modified αvβ3 integrin, results in enhanced functional activity of the integrin.40,41 Moreover, further studies indicate that cells displaying the unconventional αvβ3 integrin and cells treated with the furin inhibitor dec-RVKR-cmk show an increased attachment to type I collagen.42 These results are consistent with furin-processed αv controlling the cross-talk between αvβ3 and α2β1 integrins. Along with our own results, this clearly demonstrates that the cleavage of αv subunit and the way it is cleaved can drastically influence the integrin function and consequently the behavior of malignant cells.

In addition, it is worth noting that both PDX0 and PDX39P cells synthesize vitronectin in vivo and that this protein was mostly extracellular in PDX39P-induced tumors, while it remains largely cytoplasmic in PDX0-induced tumors (not shown). Although alternative explanations are possible, we can postulate that extracellular vitronectin may facilitate the invasion process of PDX39P cells. It has been shown that vitronectin is produced, on cleavage by furin, under two functionally different forms43 and that intact vitronectin induces matrix metalloproteinase expression and enhanced melanoma cells migration and invasion.44 Accordingly, although this question has not been addressed in this study, it is possible that the expression of α1-PDX would impinge on the vitronectin cleavage and facilitate cell invasiveness in vivo.

Correlation between changes in the function or in the repertoire of integrins and alterations in cell motility or invasiveness has been observed both in vitro and in physiopathological diseases.45,46 For example, FG carcinoma cells which express the αvβ5 integrin bind to, but do not migrate on vitronectin, while variants expressing the αvβ3 integrin were able to migrate on vitronectin via an αvβ3-dependent mechanism.47 However, addition of epidermal growth factor (EGF) to parental αvβ3-deficient cells provided a PKC-dependent signal that allowed FG cells to migrate on vitronectin via an αvβ5-dependent mechanism.48 This example also shows that interactions with microenvironment (the presence of EGF in this case) may be critical for tumor cell behavior.

The tumor microenvironment modulates growth, spreading, and migration of xenografted cancer cells by secreted growth factors.3–5 We thus observed that co-culture with rat myofibroblasts significantly increased PDX39P proliferation. Migration, adhesion, and spreading were also altered by myofibroblasts-secreted factors, including HGF and uPA (see Figure 3 and data not shown). Furthermore, it was shown that the convertase expression confers an estrogen dependency for proliferation, both in vitro and in vivo, on human breast cancer cells.49 Differences in stromal microenvironment might thus explain the discrepancies between the present in vivo results and invasion assays using tracheal xenotransplants.22,25

Given the role of PCs in biological activation of inactive precursors, their inhibition could be an interesting clinical strategy against cancers by simultaneously disrupting the function of numerous proteins involved in tumor cell invasion and tumor progression.15,20 However, the effect of α1-PDX on certain convertase substrates, like adhesion molecules, MMPs or other proteases, that could potentially contribute to tumorigenesis should be taken into account when considering PC inhibitors as therapeutic tools for a variety of diseases, including cancer. Thus, inhibition of cadherin cleavage results in decreased intercellular adhesive strength,50 a situation observed in many tumors. Our present data clearly show that inhibition of PCs can also result in a dramatic increase of tumor cells aggressiveness. Whether this enhanced malignancy is specific to colonic tumor cells remains to be determined.

Thus, it appears that inhibiting PCs actively contribute to the malignant phenotype of human colonic cancer cells when injected to immunosuppressed rats. Our observations collectively suggest that the enhanced in vivo invasiveness of α1-PDX-expressing cells might be linked to the increased motility, supported by αv-containing integrins, although other molecules might also be involved.

Acknowledgments

We thank Drs. J-C. Gevrey and A. Prat for help with this study. We also thank Dr. G. Gouysse, J. Secchi, and L. Wickham for their expert technical assistance.

Footnotes

Address reprint requests to José Luis, CNRS UMR6032, Faculté de Pharmacie, 27 Bd J. Moulin, 13 385 Marseille Cedex 5, France. E-mail: jose.luis@pharmacie.univ-mrs.fr.

Supported in part by grants from the ARC (Association pour la Recherche sur le Cancer), the GEFLUC (Groupement des Entreprises Françaises dans la Lutte contre le Cancer) the Ligue Nationale contre le Cancer and the INSERM (Institut National de Santé et de Recherche Médicale), and by a CIHR grant MGP-44363 to N.G.S.

M.N. and V.B. contributed equally to this work

References

  1. Small JV, Kaverina I, Krylyshkina O, Rottner K. Cytoskeleton cross-talk during cell motility. FEBS Lett. 1999;452:96–99. doi: 10.1016/s0014-5793(99)00530-x. [DOI] [PubMed] [Google Scholar]
  2. Geiger B, Bershadsky A, Pankov R, Yamada KM. Transmembrane crosstalk between the extracellular matrix-cytoskeleton crosstalk. Nat Rev Mol Cell Biol. 2001;2:793–805. doi: 10.1038/35099066. [DOI] [PubMed] [Google Scholar]
  3. Aplin A, Howe A, Alahari S, Juliano R. Signal transduction and signal modulation by cell adhesion receptors: the role of integrins, cadherins, immunoglobulin-cell adhesion molecules, and selectins. Pharmacol Rev. 1998;50:197–263. [PubMed] [Google Scholar]
  4. Giancotti FG, Ruoslahti E. Integrin signaling. Science. 1999;285:1028–1032. doi: 10.1126/science.285.5430.1028. [DOI] [PubMed] [Google Scholar]
  5. Parise LV, Lee J, Juliano RL. New aspects of integrin signaling in cancer. Semin Cancer Biol. 2000;10:407–414. doi: 10.1006/scbi.2000.0337. [DOI] [PubMed] [Google Scholar]
  6. Clezardin P. Recent insights into the role of integrins in cancer metastasis. Cell Mol Life Sci. 1998;54:541–548. doi: 10.1007/s000180050182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Hynes RO. Cell adhesion: old and new questions. Trends Cell Biol. 1999;9:M33–M37. [PubMed] [Google Scholar]
  8. Lehmann M, Rigot V, Seidah NG, Marvaldi J, Lissitzky JC. Lack of integrin α-chain endoproteolytic cleavage in furin-deficient human colon adenocarcinoma cells LoVo. Biochem J. 1996;317:803–809. doi: 10.1042/bj3170803. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Lissitzky JC, Luis J, Munzer JS, Benjannet S, Parat F, Chrétien M, Marvaldi J, Seidah NG. Endoproteolytic processing of integrin pro-α subunits involves the redundant function of furin and proprotein convertase (PC) 5A, but not paired basic amino acid converting enzyme (PACE) 4, PC5B, or PC7. Biochem J. 2000;346:133–138. [PMC free article] [PubMed] [Google Scholar]
  10. Berthet V, Rigot V, Champion S, Secchi J, Fouchier F, Marvaldi J, Luis J. Role of endoproteolytic processing in the adhesive and signaling functions of αvβ5 integrin. J Biol Chem. 2000;275:33308–33313. doi: 10.1074/jbc.M004834200. [DOI] [PubMed] [Google Scholar]
  11. Kolodziej M, Vilaire G, Gonder D, Poncz M, Bennett J. Study of the endoproteolytic cleavage of platelet glycoprotein IIb using oligonucleotide-mediated mutagenesis. J Biol Chem. 1991;266:23499–23504. [PubMed] [Google Scholar]
  12. Teixidó J, Parker CM, Kassner PD, Hemler ME. Functional and structural analysis of VLA-4 integrin α4 subunit cleavage. J Biol Chem. 1992;267:1786–1791. [PubMed] [Google Scholar]
  13. Delwel GO, Hogervorst F, Sonnenberg A. Cleavage of the α6A subunit is essential for activation of the α6Aβ1 integrin by phorbol 12-myristate 13-acetate. J Biol Chem. 1996;271:7293–7296. doi: 10.1074/jbc.271.13.7293. [DOI] [PubMed] [Google Scholar]
  14. Walker JL, Zhang L, Menko AS. A signaling sole for the uncleaved form of α6 integrin in differentiating lens fiber cells. Dev Biol. 2002;251:195–205. doi: 10.1006/dbio.2002.0823. [DOI] [PubMed] [Google Scholar]
  15. Taylor NA, Van De Ven WJ, Creemers JW. Curbing activation: proprotein convertases in homeostasis and pathology. EMBO J. 2003;17:1215–1227. doi: 10.1096/fj.02-0831rev. [DOI] [PubMed] [Google Scholar]
  16. Jean F, Stella K, Thomas L, Liu G, Xiang Y, Reason A, Thomas G. α1-antitrypsin Portland, a bioengineered serpin highly selective for furin: application as an antipathogenic agent. Proc Natl Acad Sci USA. 1998;95:7293–7298. doi: 10.1073/pnas.95.13.7293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Denault J-B, D’Orléans-Juste P, Masaki T, Leduc R. Inhibition of proconvertase-related processing of proendothelin-1. J Cardiovasc Pharmacol. 1995;26:S47–S50. [PubMed] [Google Scholar]
  18. Bennett BD, Denis P, Haniu M, Teplow DB, Kahn S, Louis JC, Citron M, Vassar R. A furin-like convertase mediates propeptide cleavage of BACE, the Alzheimer’s β-secretase. J Biol Chem. 2000;275:37712–37717. doi: 10.1074/jbc.M005339200. [DOI] [PubMed] [Google Scholar]
  19. Benjannet S, Elagoz A, Wickham L, Mamarbachi M, Munzer JS, Basak A, Lazure C, Cromlish JA, Sisodia S, Checler F, Chrétien M, Seidah NG. Post-translational processing of β-secretase (β-amyloid-converting enzyme) and its ectodomain shedding: the pro- and transmembrane/cytosolic domains affect its cellular activity and amyloid-β production. J Biol Chem. 2001;276:10879–10887. doi: 10.1074/jbc.M009899200. [DOI] [PubMed] [Google Scholar]
  20. Khatib AM, Siegfried G, Chrétien M, Metrakos P, Seidah NG. Proprotein convertases in tumor progression and malignancy: novel targets in cancer therapy. Am J Pathol. 2002;160:1921–1935. doi: 10.1016/S0002-9440(10)61140-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Bassi DE, Mahloogi H, Al-Saleem L, Lopez De Cicco R, Ridge JA, Klein-Szanto AJ. Elevated furin expression in aggressive human head and neck tumors and tumor cell lines. Mol Carcinog. 2001;31:224–232. doi: 10.1002/mc.1057. [DOI] [PubMed] [Google Scholar]
  22. Bassi DE, Mahloogi H, Lopez De Cicco R, Klein-Szanto A. Increased furin activity enhances the malignant phenotype of human head and neck cancer cells. Am J Pathol. 2003;162:439–447. doi: 10.1016/s0002-9440(10)63838-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Khatib AM, Siegfried G, Prat A, Luis J, Chrétien M, Metrakos P, Seidah NG. Inhibition of proprotein convertases is associated with loss of growth and tumorigenicity of HT-29 human colon carcinoma cells: importance of insulin-like growth factor-1 (IGF-1) receptor processing in IGF-1-mediated functions. J Biol Chem. 2001;276:30686–30693. doi: 10.1074/jbc.M101725200. [DOI] [PubMed] [Google Scholar]
  24. Mercapide J, Lopez De Cicco R, Bassi DE, Castresana JS, Thomas G, Klein-Szanto AJ. Inhibition of furin-mediated processing results in suppression of astrocytoma cell growth and invasiveness. Clin Cancer Res. 2002;8:1740–1746. [PubMed] [Google Scholar]
  25. Bassi DE, Lopez De Cicco R, Mahloogi H, Zucker S, Thomas G, Klein-Szanto AJ. Furin inhibition results in absent or decreased invasiveness and tumorigenicity of human cancer cells. Proc Natl Acad Sci USA. 2001;98:10326–10331. doi: 10.1073/pnas.191199198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Timpl R, Rohde H, Gehron-Robey P, Rennard SI, Foidart JM, Martin GR. Laminin: a glycoprotein from basement membranes. J Biol Chem. 1979;254:9933–9937. [PubMed] [Google Scholar]
  27. Yatogho T, Izumi M, Kashiwagi H, Hayashi M. Novel purification of vitronectin from human plasma by heparin affinity chromatography. Cell Struct Funct. 1988;13:281–292. doi: 10.1247/csf.13.281. [DOI] [PubMed] [Google Scholar]
  28. Lehmann M, Rabenandrasana C, Tamura R, Lissitzky J-C, Quaranta V, Pichon J, Marvaldi J. A monoclonal antibody inhibits adhesion to fibronectin and vitronectin of a colon carcinoma cell line and recognizes the integrins αvβ3, αvβ5, and αvβ6. Cancer Res. 1994;54:2102–2107. [PubMed] [Google Scholar]
  29. Carreiras F, Rigot V, Cruet S, André F, Gauduchon P, Marvaldi J. Migration properties of the human ovarian adenocarcinoma cell line IGROV1: importance of αvβ3 integrins and vitronectin. Int J Cancer. 1999;80:285–294. doi: 10.1002/(sici)1097-0215(19990118)80:2<285::aid-ijc19>3.0.co;2-l. [DOI] [PubMed] [Google Scholar]
  30. Rigot V, Lehmann M, André F, Daemi N, Marvaldi J, Luis J. Integrin ligation and PKC activation are required for migration of colon carcinoma cells. J Cell Sci. 1998;111:3119–3127. doi: 10.1242/jcs.111.20.3119. [DOI] [PubMed] [Google Scholar]
  31. Vakaet L, Jr, Vleminckx K, Van Roy F, Mareel M. Numerical evaluation of the invasion of closely related cell lines into collagen type I gels. Invasion Metastasis. 1991;11:249–260. [PubMed] [Google Scholar]
  32. Dore JF, Bailly M, Bertrand S. Metastases of human tumors in experimental animals. Anticancer Res. 1987;7:997–1003. [PubMed] [Google Scholar]
  33. André F, Rigot V, Remacle-Bonnet M, Luis J, Pommier G, Marvaldi J. Protein kinases C-γ and -δ are involved in insulin-like growth factor I-induced migration of colonic epithelial cells. Gastroenterology. 1999;116:64–77. doi: 10.1016/s0016-5085(99)70230-1. [DOI] [PubMed] [Google Scholar]
  34. Komada M, Hatsuzawa K, Shibamoto S, Ito F, Nakayama K, Kitamura N. Proteolytic processing of the hepatocyte growth factor/scatter factor receptor by furin. FEBS Lett. 1993;328:25–29. doi: 10.1016/0014-5793(93)80958-w. [DOI] [PubMed] [Google Scholar]
  35. Daemi N, Thomasset N, Lissitzky JC, Dumortier J, Jacquier MF, Pourreyron C, Rousselle P, Chayvialle JA, Rémy L. Anti-β4 integrin antibodies enhance migratory and invasive abilities of human colon adenocarcinoma cells and their MMP-2 expression. Int J Cancer. 2000;85:850–856. doi: 10.1002/(sici)1097-0215(20000315)85:6<850::aid-ijc19>3.0.co;2-b. [DOI] [PubMed] [Google Scholar]
  36. Dumortier J, Ratineau C, Scoazec JY, Pourreyron C, Anderson W, Jacquier MF, Blanc M, Bernard C, Bellaton C, Rémy L, Chayvialle JA, Roche C. Site-specific epithelial-mesenchymal interactions in digestive neuroendocrine tumors: an experimental in vivo and in vitro study. Am J Pathol. 2000;156:671–683. doi: 10.1016/S0002-9440(10)64771-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Berthet V, Rigot V, Nejjari, Marvaldi J, Luis J. The endoproteolytic processing of αvβ5 integrin is involved in cytoskeleton remodelling and cell migration. FEBS Lett. 2004;557:159–163. doi: 10.1016/s0014-5793(03)01467-4. [DOI] [PubMed] [Google Scholar]
  38. Holly SP, Larson MK, Parise LV. Multiple roles of integrins in cell motility. Exp Cell Res. 2000;261:69–74. doi: 10.1006/excr.2000.5040. [DOI] [PubMed] [Google Scholar]
  39. Palecek SP, Loftus JC, Ginsberg MH, Lauffenburger DA, Horwitz AF. Integrin-ligand binding properties govern cell migration speed through cell-substratum adhesiveness. Nature. 1997;385:537–540. doi: 10.1038/385537a0. [DOI] [PubMed] [Google Scholar]
  40. Deryugina EI, Ratnikov BI, Postnova TI, Rozanov DV, Strongin AY. Processing of integrin αv subunit by MT1-MMP stimulates migration of breast carcinoma cells on vitronectin and enhances tyrosine phosphorylation of FAK. J Biol Chem. 2002;277:9749–9756. doi: 10.1074/jbc.M110269200. [DOI] [PubMed] [Google Scholar]
  41. Ratnikov BI, Rozanov DV, Postnova TI, Baciu PG, Zhang H, DiScipio RG, Chestukhina GG, Smith JW, Deryugina EI, Strongin AY. An alternative processing of integrin αv subunit in tumor cells by membrane type-1 matrix metalloproteinase. J Biol Chem. 2002;277:7377–7385. doi: 10.1074/jbc.M109580200. [DOI] [PubMed] [Google Scholar]
  42. Baciu PC, Suleiman EA, Deryugina EI, Strongin AY. Membrane type-1 matrix metalloproteinase (MT1-MMP) processing of pro-αv integrin regulates cross-talk between αvβ3 and α2β1 integrins in breast carcinoma cells. Exp Cell Res. 2003;291:167–175. doi: 10.1016/s0014-4827(03)00387-2. [DOI] [PubMed] [Google Scholar]
  43. Seger D, Shaltiel S. Evidence showing that the two-chain form of vitronectin is produced in the liver by a selective furin cleavage. FEBS Lett. 2000;480:169–174. doi: 10.1016/s0014-5793(00)01917-7. [DOI] [PubMed] [Google Scholar]
  44. Bafetti LM, Young TN, Itoh Y, Stack MS. Intact vitronectin induces matrix metalloproteinase-2 and tissue inhibitor of metalloproteinases-2 expression and enhanced cellular invasion by melanoma cells. J Biol Chem. 1998;273:143–149. doi: 10.1074/jbc.273.1.143. [DOI] [PubMed] [Google Scholar]
  45. Couvelard A, Leseche G, Scoazec JY, Groussard O. Human allograft vein failure: immunohistochemical arguments supporting the involvement of an immune-mediated mechanism. Hum Pathol. 1995;26:1313–1320. doi: 10.1016/0046-8177(95)90295-3. [DOI] [PubMed] [Google Scholar]
  46. Nejjari M, Hafdi Z, Dumortier J, Bringuier AF, Feldmann G, Scoazec JY. α6β1 integrin expression in hepatocarcinoma cells: regulation and role in cell adhesion and migration. Int J Cancer. 1999;83:518–525. doi: 10.1002/(sici)1097-0215(19991112)83:4<518::aid-ijc14>3.0.co;2-q. [DOI] [PubMed] [Google Scholar]
  47. Leavesley DI, Ferguson GD, Wayner EA, Cheresh DA. Requirement of the integrin β3 subunit for carcinoma cell spreading or migration on vitronectin and fibrinogen. J Cell Biol. 1992;117:1101–1107. doi: 10.1083/jcb.117.5.1101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Klemke RL, Yebra M, Bayna EM, Cheresh DA. Receptor tyrosine kinase signaling required for integrin αvβ5-directed cell motility but not adhesion on vitronectin. J Cell Biol. 1994;127:859–866. doi: 10.1083/jcb.127.3.859. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Cheng M, Xu N, Iwasiow B, Seidah N, Chrétien M, Shiu RP. Elevated expression of proprotein convertases alters breast cancer cell growth in response to estrogen and tamoxifen. J Mol Endocrinol. 2001;26:95–105. doi: 10.1677/jme.0.0260095. [DOI] [PubMed] [Google Scholar]
  50. Posthaus H, Dubois CM, Muller E. Novel insights into cadherin processing by subtilisin-like convertases. FEBS Lett. 2003;536:203–208. doi: 10.1016/s0014-5793(02)03897-8. [DOI] [PubMed] [Google Scholar]

Articles from The American Journal of Pathology are provided here courtesy of American Society for Investigative Pathology

RESOURCES