Autocrine induction of invasion and metastasis by tumor-associated trypsin inhibitor in human colon cancer cells


From the conditioned medium of the human colon carcinoma cells, HT-29 5M21 (CM-5M21), expressing a spontaneous invasive phenotype, tumor-associated trypsin inhibitor (TATI) was identified and characterized by proteomics, cDNA microarray approaches and functional analyses. Both CM-5M21 and recombinant TATI, but not the K18Y-TATI mutant at the protease inhibitor site, trigger collagen type I invasion by several human adenoma and carcinoma cells of the colon and breast, through phosphoinositide-3-kinase, protein kinase C and Rho-GTPases/Rho kinase-dependent pathways. Conversely, the proinvasive action of TATI in parental HT29 cells was alleviated by the TATI antibody PSKAN2 and the K18Y-TATI mutant. Stable expression of K18Y-TATI in HT-29 5M21 cells downregulated tumor growth, angiogenesis and the expression of several metastasis-related genes, including CSPG4 (13.8-fold), BMP-7 (9.7-fold), the BMP antagonist CHORDIN (5.2-fold), IGFBP-2 and IGF2 (9.6- and 4.6-fold). Accordingly, ectopic expression of KY-TATI inhibited the development of lung metastases from HT-29 5M21 tumor xenografts in immunodeficient mice. These findings identify TATI as an autocrine transforming factor potentially involved in early and late events of colon cancer progression, including local invasion of the primary tumor and its metastatic spread. Targeting TATI, its molecular partners and effectors may bring novel therapeutic applications for high-grade human solid tumors in the digestive and urogenital systems.


Tumor-associated trypsin inhibitor (TATI) is a 56 amino-acid peptide (molecular weight=6.2 kDa). TATI is the serine protease inhibitor Kazal type 1 (SPINK1) and is also called pancreatic secretory inhibitor (PSTI) since this peptide is secreted together with trypsinogen by pancreatic acinar cells and in the gastrointestinal and urinary tract (Kazal et al., 1948; Paju and Stenman, 2006). TATI inhibits trypsin with strong affinity, as well as plasmin and urokinase, however with lower affinities (Turpeinen et al., 1988). The interaction site has been identified as lysine-18 and isoleucine-19 within the N-terminal part of TATI (Bartelt et al., 1977). This inhibition is reversible since the complex is dissociated by serum: trypsin is released from TATI and complexed with α2-macroglobulin and α1-protease inhibitor. TATI is also inactivated by trypsin cleavage between Arg 42-Lys 43 and Arg 44-Gln 45 (Kikuchi et al., 1989). TATI has been originally found in urine of a patient with ovarian cancer (Huhtala et al., 1982). Later, high levels of TATI in tumors and/or body fluids were reported in cancers of pancreas, colon, liver, lung and breast (Paju and Stenman, 2006).

To our knowledge, the possible contribution of TATI in the molecular and cellular mechanisms involved in cancer progression are unknown. We have previously reported that the mucin-secreting cell clone (HT-29 5M21) isolated from a subpopulation of human colon cancer cells HT-29 cells resistant to methotrexate (Lesuffleur et al., 1998) displayed spontaneous cellular self-aggregation and invasive phenotypes in type I collagen gels (Truant et al., 2003). We, therefore, investigated the molecular and cellular mechanisms underlying this aggressive behavior. The results provide the first evidence for TATI as a major autocrine and transforming factor produced by HT-29 5M21 cells to control colon cancer cell invasion and metastasis.


HT-29 5M21 cells secrete autocrine proinvasive factor(s)

We have previously shown that the HT-29-derived subclone 5M21 displayed a constitutive invasive behavior in type I collagen whereas the parental HT-29 STD cells were not spontaneously invasive in this extracellular matrix component (Truant et al., 2003). In order to verify the possibility that this spontaneous invasive behavior could be mediated by autocrine proinvasive factor(s) secreted by HT-29 5M21 cells, we have analysed the proinvasive potential of the conditioned medium (CM) prepared from these cells. We used parental HT-29 STD and HCT-8/S11 colon cancer cells, respectively, isolated from LOH and MSI sporadic tumors, as well as premalignant adenoma PC/AA/C1 cells derived from a patient with familial adenomatous polyposis, and their src-transformed counterparts, the PC/msrc cells (Empereur et al., 1997). As shown in Figure 1, addition of the CM from HT-29 5M21 cells induced the invasion of collagen type I gels by these four cell lines whereas the CM prepared from HT-29 STD cells was ineffective. This finding encouraged us to further characterize the presence of autocrine proinvasive factor(s) produced by HT-29 5M21 cells.

Figure 1

HT-29 5M21 conditioned medium (CM) induces collagen type I invasion by premalignant and malignant human colon epithelial cells. Collagen invasion assays were performed in control conditions (no addition, open bars) and in the presence of the CM prepared from either HT-29 5M21 cells (hatched bar) or HT-29 STD cells (filled bar). Data are mean±s.e.m. from 2–3 experiments.

TATI is overexpressed and secreted by HT-29 5M21 cells

Next, we compared the protein profiles in the CM prepared from HT-29 5M21 and parental HT-29 STD cells. Proteins were separated by ultrafiltration according to their molecular weight (MW), using membranes at a cutoff of 3–10 kDa. Protein fractions were assayed for proinvasive activity in vitro. We observed that the proinvasive activity of the HT-29 5M21 CM was only present in protein fractions ranging from 3 to 10 kDa (not shown). Comparative proteomics was then undertaken on the 3–10 kDa fractions prepared from HT-29 5M21 and HT-29 STD CM. The first approach was a separation by 2D-electrophoresis in conditions adapted for low molecular mass proteins, followed by analysis of tryptic peptides by MALDI-TOF mass spectrometry. Regarding the low resolution in 2D-gels, direct analysis was performed, followed by protein separation by high performance liquid chromatography (HPLC), and N-terminal sequencing. As shown in Figure 2a, the MALDI spectra identified TATI (m/z, 6245.4) in HT-29 5M21 CM but not in HT-29 STD CM (Figure 2b). Transcriptome analysis using HT-29 5M21 and STD cells cultured on type I collagen was then performed using the 44K Whole Human Genome Oligo Microarray (Agilent Technologies, Santa Clara, CA, USA). Data were normalized and processed using Genespring Software (Agilent Technologies, Palo Alto, USA) to select upregulated genes. Table 1 shows the list of genes that are upregulated more than 10-fold in HT-29 5M21 cells versus HT-29 STD cells. The pangenomic microarray analysis showed that TATI (SPINK1) was one of the most strongly upregulated genes in HT-29 5M21 cells (expression ratio of about 14), versus parental cells. This data was confirmed by real-time RT–PCR (expression ratio=16). SPINK 1 is a Kazal type 1 serine protease inhibitor, thus not a priori a likely proinvasive agent. However, its presence in the CM harvested from the HT-29 5M21 cells justified further experiments to address its function in invasion. Other genes highly expressed by HT-29 5M21 cells are critically involved in cancer progression. For instance, the chemokine (C-X-C motif) receptor 4 (CXCR4) is involved at different levels in the regulation of malignancy through its interaction with its chemokine ligand stromal cell-derived factor-1 (SDF-1/CXCL12). CXCR4 promotes the metastatic spread of cancer cells to distant organs expressing SDF-1/CXCL12 and is associated with poor patient survival. The production of SDF-1/CXCL12 by stromal cancer-associated fibroblasts promotes tumor cell growth and stimulates tumor angiogenesis by recruiting endothelial progenitor cells (Ben-Baruch, 2006; Burger and Kipps, 2006). Growth factor receptor-bound protein 10 (Grb10) is an adaptor protein interacting with several cell survival-related binding partners, notably Raf-1 kinase in a Bad-dependent manner (Kebache et al., 2007). Among the three enzymes highly upregulated in HT-29 5M21 cells versus control HT-29 STD cells, that is, creatine kinase B (CKB) and aldo-keto reductase family 1 (AKRB1) B1 and B10 members, two of them are associated with colon cancer: CKB showed an altered expression and an abnormal localization in the nuclear matrix (Balasubramani et al., 2006); AKR1B10 was shown to stimulate the proliferation of colon cancer cells (Yan et al., 2007).

Figure 2

TATI is the major proinvasive factor secreted by HT-29 5M21 cells. (a, b) MALDI-TOF mass spectrometry showing that TATI is present in the conditioned medium (CM) prepared from HT-29 5M21 cells (a), but absent in the CM of HT-29 STD cells (b). (c) Collagen type I invasion assays were performed using the invasive HT-29 5M21 cells cultured in the absence (control conditions, gray bar) and presence of the TATI antibody PSKAN2 (filled bar). Data were compared with HT-29 STD cells cultured in the presence (hatched bar) or absence (control conditions, open bar) of the HT-29 5M21 CM, either alone or combined with the TATI antibody (dotted bar).

Table 1 Upregulated genes (ratio>10, n=10) in HT-29 5M21 cells versus parental HT-29 STD human colon cancer cells, and their corresponding GenBank accession numbers

TATI is a proinvasive factor

Several experiments were undertaken to test the possible action of TATI in cancer cell invasion. First, we showed that the TATI antibody PSKAN2 at a dilution of 1/50 abolished the spontaneous invasive behavior of HT-29 5M21 cells, as well as the proinvasive response of HT-29 STD cells induced by the 5M21 CM (Figure 2c). Next, we examined the proinvasive activity of recombinant wild-type TATI (WT-TATI) and mutant K18Y-TATI encoded by the bacterial expression vector pQE-9 (Figure 3a). The corresponding recombinant proteins (MW=6 kDa) were identified by western blot, as shown in Figure 3a. Invasion assays were carried out using various concentrations of recombinant WT-TATI (Figure 3b). Of note, WT-TATI dose-dependently induced the invasive potential of HT-29 STD cells at the threshold concentration of 5 ng ml−1, up to 16 ng ml−1, followed by a progressive decline at higher concentrations. Then, the invasive activities of both WT-TATI and KY-TATI was analysed at the optimal concentration of 16 ng ml−1. As shown in Figure 3c, recombinant WT-TATI, but not recombinant KY-TATI, induced the invasive behavior of human adenoma and carcinoma cells (HT-29 STD, HCT8/S11, PC/AAC1, PC/msrc), and MCF-7 breast cancer cells as well. In addition, recombinant KY-TATI dose-dependently reversed the proinvasive effect of recombinant WT-TATI on HT-29 STD cells (Figure 3d). We also examined the impact of WT-TATI and K18Y-TATI ectopic expression in HT-29 STD cells. For this purpose, HT-29 STD cells were stably transfected with the pcDNA3.1 expression vector encoding either WT-TATI or K18Y-TATI under the hybrid promoter EF1α-HTLV. Ectopic expression of WT-TATI induced collagen type I invasion by HT-29 STD cells (invasion index=11.5%) while stable expression of K18Y-TATI was ineffective (invasion index=1.8%).

Figure 3

Wild-type recombinant TATI (WT-TATI) is a proinvasive factor. (a) Procaryote expression vectors (pQE-9) were constructed to produce N-His-tag WT (1) and mutated K18Y* (2) TATI from human cDNA (3). The star indicates the single amino-acid mutation (K18Y*) introduced by site-directed mutagenesis. The recombinant proteins (WT-TATI: WT; mutated-TATI: KY, versus the negative control C: empty plasmid pQE-9) were produced and lysates were resolved by 15% SDS–PAGE. Proteins were detected via Ponceau staining (1) or immunoblotting (2) with the anti-TATI antibody; M: molecular weight marker (SeeBlue 2; Invitrogen). (b) Dose effect of recombinant WT-TATI on collagen type I invasion by HT-29 STD cells. The optimal concentration of WT-TATI was 16 ng ml−1. (c) Recombinant WT-TATI is a proinvasive factor for human colon and breast cancer cells. Invasion assays were performed with HT-29 STD, HCT-8/S11, PC/AA/C1, PC/msrc and MCF-7/AZ cells in control conditions (open bars) and in the presence of recombinant WT-TATI (16 ng ml, filled bar), either alone or combined with the anti-TATI antibody (hatched bar), or recombinant KY-TATI (16 ng ml−1, dotted bar). (d) Antagonism between WT-TATI (16 ng ml−1) and increasing concentrations of recombinant mutated KY-TATI at collagen type I invasion by HT-29 STD cells. Data are mean±s.e.m. from 2–3 experiments.

The proinvasive activity of TATI is connected with several oncogenic pathways involved in cancer cell invasion and metastasis

Cell migration is controlled by the small GTPases RhoA, Rac1 and Cdc42 that regulate actin dynamics and promote the assembly of integrin-based matrix adhesion complex (Raftopoulou and Hall, 2004). In addition, PI3-kinase and several Protein Kinases C (PKC) isoforms are well characterized as critical effectors of multiple invasion pathways induced by oncogenes, hormones and growth factors (Kotelevets et al., 1998). Accordingly, the Rho-GTPase inhibitor C3 T exoenzyme, and the Rho-kinase (ROCK) inhibitor Y27632 abolished TATI-induced invasion by HT-29 STD cells (Figure 4a). Moreover, function-blocking mutants of the Rho-GTPases RhoA (N19), Cdc42 (N17) and Rac1 (N17) stably transfected in human colon cancer cells HCT8/S11 (Nguyen et al., 2005) prevented the TATI invasive activity (Figure 4b). Pharmacological inhibitors targeting PI3-kinase (wortmaninn) and PKC (GF109 and Gö6976) obliterated the proinvasive action of recombinant TATI in HT-29 STD cells. In contrast, the proinvasive action of TATI was unaffected by pharmacological blockade of the mitogen-activated protein kinases (MAPKs) p38 (SB203580) and p42/p44 (PD098059).

Figure 4

Implication of several proinvasive pathways in TATI-induced colon cancer cell invasion. (a) Collagen type I invasion assays were performed using HT-29 STD cells in control conditions (open bar) and in the presence of recombinant WT-TATI (16 ng ml−1), either alone or combined with pharmacological inhibitors targeting PI3-kinase (wortmaninn, 10 nM), Rho-GTPases (C3 T exoenzyme, 5 μg ml−1), ROCK (Y27632, 1 μM), PKCs (Gö6976, 1 μM and GF109, 1 μM), p38 (SB203580, 10 μM) and p42/44 (PD098059, 50 μM) MAPKs. (b) Invasion assays were performed using HCT-8/S11 cells stably transfected with dominant negative forms of the Rho-GTPAses (RhoA-T19N, Cdc42-T17N and Rac1-T17N), and cultured in control conditions (open bars) and following the addition of recombinant WT-TATI (16 ng ml−1, filled bars).

KY-TATI inhibits the growth and metastatic dissemination of HT-29 5M21 tumor xenografts

In order to investigate the contribution of TATI to the growth and dissemination of colon cancer cell xenografts, ectopic expression of the KY-TATI mutant was performed in HT-29 5M21 cells. When cells were injected subcutaneously in the flank of the SCID mice, the median volume of the tumors after 41 days was 1437 mm3 for the control 5M21 xenografts and 747 mm3 for the KY-TATI xenografts (P=0.012), as shown in Figure 5a. The tumor doubling time (200–400 mm3) was 10.7 days in the control group 5M21 and 15.3 days in KY-TATI xenografts. At the histological level, no difference was observed between the primary subcutaneous tumors developed by the two cell types (Figure 5b). Tumors appeared as carcinomas with cell rows and glands containing mucus in their lumen. The evaluation of tumor angiogenesis by collagen IV labeling showed a decrease in the number and size of the blood vessels in KY-TATI xenografts, in comparison to HT-29 5M21 xenografts. In contrast, the Ki-67 cell proliferation index for cells engaged in the cell cycle was found at comparable high levels in both HT-29 5M21 and KY-TATI xenografts (91 and 94%, respectively). Direct assessment of cell counts showed that the KY-TATI mutant reduced the proliferation of HT29 5M21 cells in culture at days 4–5 (25–28% inhibition) and day 7 (40% inhibition), data not shown. These findings suggest that the KY-TATI mutant targets tumor growth and angiogenesis defects by compromising tumor take and cell survival in the KY-TATI-expressing colon HT-29 5M21 xenografts. It is therefore plausible to consider TATI as a survival factor in a variety of human solid tumors engaged in the mucinous differentiation phenotype. We found that HT-29 5M21 tumors developed metastases in the lungs only. Therefore, the lungs were systematically examined by histological examination for the detection of metastases (10 sections per lung in the two groups of mice), in three separate experiments. Metastases and/or micrometastases in lungs (black arrows) contained mucus-positive cells and were observed in 8/11 mice bearing control 5M21 xenografts and in only 3/12 mice for the KY-TATI group (Figure 5b). For the control HT-29 5M21 xenografts, their associated metastases contained numerous cancer cells with mucin secretory activity, whereas the KY-TATI group showed micrometastases with scarce and rare cancer cells. Thus, ectopic expression of KY-TATI inhibited the formation of lung metastases following the implantation of subcutaneous HT-29 5M21 xenografts in SCID mice. In order to identify the mechanisms underlying the antimetastatic function of KY-TATI, we compared the gene expression patterns in KY-TATI and control HT-29 5M21 tumor xenografts at 41 days postinoculation of the cancer cells in SCID mice. Transcriptome analysis of the corresponding primary tumors was performed using the 44K Whole Human Genome Oligo Microarray (Agilent Technologies). Data were normalized and processed using the Genespring Software (Silicon Genetics) to select the regulated genes using the expression value cutoff of 2 (828 genes). Then, these selected genes were further sorted through the Ingenuity Pathways Analysis System ( to identify metastasis-associated genes (28 genes, Table 2). Data in Table 2 showed that the majority of these genes were downregulated by KY-TATI. The most critically compromised genes encode chondroitin sulfate proteoglycan 4 (CSPG4), bone morphogenetic protein 7 (BMP-7), chordin (CHRD), insulin-like growth factor binding protein 2 (IGFBP2) and insulin-like growth factor (IGF2). Conversely, we found that ectopic expression of KY-TATI upregulated 8.7-fold the gene encoding the CCL2 chemokine. However, divergent data are reported on the role of CCL2 in metastasis. Our data further illustrate the versatility of chemokines and their receptors in cancer progression, in view of their pleiotropic roles on immune cells (infiltration of tumor-associated macrophages, dendritic cells) and on tumor-associated stromal cells and cancer cells. It is possible that such a versatility is also related to the proliferation potential and differentiation traits of epithelial cancers, in the context of the neuroendocrine or mucinous lineages; see also Table 2 legend for other comments.

Figure 5

Impact of the ectopic expression of the KY-TATI mutant on the growth and dissemination of human colon cancer HT-29 5M21 xenografts in immunodeficient mice. (A) Control HT-29 5M21 cells and their KY-TATI counterparts stably transfected by the mutant TATI at the protease interacting site were injected subcutaneously as xenografts in SCID mice (2 × 106 cells). Tumor growth was monitored for 41 days, twice a week. Data are from 3–4 animals for each group and are representative of two other experiments. Significant differences at *P=0.05, #P=0.025, **P=0.012. (B) Histological (hematein eosin saffron astra blue: HESAB, magnification: × 400) and immunohistochemical staining of collagen IV (Col IV, magnification: × 50) of primary tumor xenografts and lung metastases (arrows) was performed at day 41 following subcutaneous injections of HT-29 5M21 and KY-TATI colon cancer cells. (a, d) Primary subcutaneous xenografts developed as adenocarcinomas with mucus in glandular lumen. (b, e) The KY-TATI mutant reduced angiogenesis (blood microvessels size and density) in KY-TATI xenografts in comparison to control HT-29 5M21 xenografts. (c, f) The presence of metastases containing colon cancer cells with mucin secretions was detected in the lungs of mice bearing control HT-29 5M21 xenografts.

Table 2 Regulated genes (ratio>2, n=28) in K18Y-TATI tumor xenografts versus HT-29 5M21 tumor xenografts at 41 days post inoculation of the cancer cells in SCID mice, and their corresponding GenBank accession numbers


Our comparative study using proteomics and microarray approaches showed that the Kazal-type inhibitor of serine proteases TATI is highly expressed and secreted by human colon cancer cells, HT-29 5M21, expressing a differentiated mucinous phenotype. This overexpression of TATI in 5M21 cells mediates a constitutive and autocrine invasive potential in vitro. Accordingly, ectopic expression of the function-blocking mutant KY-TATI at the serine–protease interaction site negatively regulates the invasive and metastatic potential of HT-29 5M21 cancer cell xenografts to the lungs. In this report, we showed for the first time that TATI is a transforming factor involved in the invasive and metastatic potential of human colon cancer cells.

An important finding in our study was the proinvasive activities of the HT-29 5M21 CM and of the recombinant WT-TATI on premalignant human adenoma colon cells PC/AA/C1. These data suggest that TATI may participate at the signal transduction mechanisms involved at the adenoma–carcinoma transition during the neoplastic progression in colon cancer patients. Of note, the TATI proinvasive activity was demonstrated in a series of transformed human colon cancer cell lines and MCF-7 breast cancer cells. Our data show that several proinvasive and oncogenic pathways, including PI3-kinase, Rho-like GTPases and PKC (Kotelevets et al., 1998; Le Floch et al., 2005; Nguyen et al., 2005), are implicated in the TATI invasion signals. Our initial data on the negative impact of the KY-TATI mutant on the neovascularization of the 5M21 tumor xenografts showed that TATI is involved in tumor angiogenesis, a critical step linked to the metastatic potential of human solid tumors. In addition, we found that the growth of the HT-29 5M21 cancer cell xenografts was reduced by 50% following the ectopic expression of the function-blocking mutant KY-TATI, suggesting that TATI is a permissive factor for tumor growth and cell survival. Taken together, our data suggest that TATI is involved, at least in part, in the regulation of tumor growth, cell survival and angiogenesis, through trypsin-independent mechanisms. Biological targets underlying TATI overexpression may include other mechanisms and substrates, such as plasmin and urokinase (Turpeinen et al., 1988).

Our data imply that TATI functions as an invasion promoter and a metastasis progression factor at the proliferation–differentiation and invasion–metastasis transitions during cancer progression. This conclusion is supported by a series of experimental and clinical observations linked to the growth and metastasis of human solid tumors in the context of the reciprocal interactions between proteases, their natural inhibitors and targets. For example, high levels of other protease inhibitors such as plasminogen activator inhibitor type-1 (PAI-1) and tissue inhibitor of matrix metalloproteinase type-1 (TIMP-1) are associated with a poor prognosis (McCarthy et al., 1999; Janicke et al., 2001; Gouyer et al., 2005; Harbeck et al., 2007). Both PAI-1 and TIMP-1 contribute to tumor development through cancer cell growth and survival, tumor angiogenesis and invasion (Bajou et al., 1998; Chirco et al., 2006). Increased levels of trypsinogen, tumor-associated trypsinogens, trypsin and its inhibitor TATI correlate with the malignancy of human solid tumors (Nyberg et al., 2006). The presence of TATI in tumors is a marker of adverse prognosis for hepatocellular carcinoma (HCC), bladder, kidney and mucinous ovarian cancers (Stenman, 2002; Antila et al., 2006; Lee et al., 2007; Paju et al., 2007). In patients with stage III and IV ovarian cancers, TATI tissue expression and elevated TATI concentration in serum were associated with adverse cancer-specific and progression-free survival (Paju et al., 2004). Of note, TATI overexpression correlates with anchorage-independent growth, high stage HCC, portal vein invasion, early tumor recurrence and metastasis (Lee et al., 2007). Our future efforts will be aimed at exploring sporadic and familial human colon tumors, as well as mucinous and nonmucinous neoplasms, for their ability to produce high levels of TATI. Indeed, increased TATI serum levels occur in 34–74% of patients with colon cancer (Stenman, 2002). The mechanisms underlying TATI overexpression in human tumors and its accumulation in the serum of cancer patients are largely unknown. This induction can be accomplished via transcriptional activation of the TATI gene promoter and epigenetic regulations, including alternative DNA methylation states. Besides two transactivating elements identified in the PSTI promoter gene, including IL-6 and AP-1 (Ohmachi et al., 1993; Yasuda et al., 1993), another regulatory sequence functions positively in pancreatic cells but as negative active element in nonpancreatic cells (Yasuda et al., 1998). In that respect, we cannot exclude the possibility that methotrexate and 5-FU-based combination chemotherapy (Louvet et al., 2000; Andre et al., 2004; Poole et al., 2006) might have the ability to upregulate the expression of the TATI gene through several mechanisms including aberrant expression of differentiation-associated phenotypes, and to influence the invasive and metastatic potential of growing tumors in a subset of breast and colon cancer patients.

In conclusion, the present study supports the idea that TATI exerts pleiotropic actions on colon cancer progression according to its cumulated impact on the growth, angiogenesis and metastatic dissemination of primary tumors. Targeting TATI and its molecular partners implicated in the invasion and metastatic cascades may bring novel therapeutic applications for a variety of human solid tumors in the digestive and urogenital systems.

Materials and methods

Proteome and DNA microarrays analysis

Proteome analysis of the CM prepared from HT-29 5M21 and STD cell cultures was performed as previously described (Delacour et al., 2005). The 2D-electrophoresis was adapted to the separation of proteins with low molecular mass (Fountoulakis et al., 1998). Identification of proteins from excised spots was carried out by analysis of the peptide mixtures from tryptic digests (Perspective Biosystems, Framingham, MA, USA), using the Voyager-DE STR device and Profound software ( In addition, CM was analysed by direct MALDI-TOF mass spectrometry. CM proteins were also separated and identified by HPLC and N-terminal sequencing. For DNA microarrays analysis, total RNA (0.5 μg) was processed and analysed on Human Whole Genome Agilent 44 K 60-mer oligonucleotide microarrays using the different Agilent kits of cRNA amplification, labeling, fragmentation, hybridization and washing according to the Two-Color Microarray-Based Gene Expression protocol (Agilent Technologies). Microarrays were scanned using an Axon GenePix 4100A scanner and the GenePix Pro extracted expression data were loaded and processed into the GeneSpring version 7.3 (Agilent Technologies) for normalization, filtering and statistic analysis. Functional analyses were generated by the Ingenuity Pathways Analysis (Ingenuity Systems,

Production of recombinant TATI proteins

The human TATI cDNA encoding WT-TATI was prepared by reverse transcription from total RNA extracted from HT-29 5M21 cells, amplified and cloned into the pCR4 vector (Invitrogen, Cergy, Pontoise, France). The K18Y mutation in the WT-TATI peptide was generated by PCR and PfuTurbo DNA polymerase (Stratagene, Amsterdam, The Netherlands). The cDNAs (WT and K18Y) were subcloned downstream of the MRGSH6 coding sequence into the BamHI/HindIII restriction enzyme sites of the pQE-9 expression vector (Qiagen, Hilden, Germany). The sequence of the WT and K18Y-TATI plasmids were sequenced to verify the integrity of the vectors. Production of the recombinant proteins from pQE-9 plasmids was carried out using M15 cells treated with isopropyl-1-thio-D-galactopyranoside (400 μM). The cell lysates were centrifuged and the supernatants were purified by affinity chromatography using a Co2+-agarose resin (Clontech, Saint-Germain-en-Laye, France). Bacterially expressed TATI was separated on 5–30% sodium dodecyl sulfate-polyacrylamide gels under reducing conditions. After transfer to a Hybond-C extra membrane (Amersham Pharmacia Biotech, Aylesbury, UK), recombinant WT and K18Y-TATI proteins were probed with the TATI mAb and revealed by the secondary anti-mouse HRP (1:4000, Sigma, St Louis, MO, USA). Membranes were washed in TBS saline containing 0.1% Tween 20 and revealed by chemiluminescence western detection (Amersham Pharmacia Biotech). The functional integrity of recombinant TATI was checked against the activity of trypsin.

Ectopic expression of TATI in HT-29 cells

The WT- and K18Y-TATI cDNAs were cloned into the eukaryotic expression vector pcDNA3.1 (Invitrogen) in which the cytomegalovirus promoter was replaced by the hybrid promoter EF1α-HTLV. HT-29 STD cells (2 × 106 per 10 mm petri dishes) were transfected for 48 h by each expression vector (2 μg of plasmid), using Effectene (Qiagen). Geneticin-resistant colonies were selected for 2 weeks in medium containing 500 μg ml−1 G418 and isolated as individual clones. HT-29 5M21 cells were stably transfected with the K18Y-TATI expression vector according to the same procedure.

See Supplementary Information for Materials, cell culture, qPCR, cellular invasion, tumor growth and metastasis.

Accession codes




  1. Andre T, Boni C, Mounedji-Boudiaf L, Navarro M, Tabernero J, Hickish T et al. (2004). Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. N Engl J Med 350: 2343–2351.

    CAS  Article  Google Scholar 

  2. Antila R, Jalkanen J, Heikinheimo O . (2006). Comparison of secondary and primary ovarian malignancies reveals differences in their pre- and perioperative characteristics. Gynecol Oncol 101: 97–101.

    Article  Google Scholar 

  3. Bajou K, Noel A, Gerard RD, Masson V, Brunner N, Holst-Hansen C et al. (1998). Absence of plasminogen activator inhibitor 1 prevents cancer invasion and vascularization. Nat Med 4: 923–928.

    CAS  Article  Google Scholar 

  4. Bartelt DC, Shapanka R, Greene LJ . (1977). The primary structure of the human pancreatic secretory trypsin inhibitor. Amino acid sequence of the reduced S-aminoethylated protein. Arch Biochem Biophys 179: 189–199.

    CAS  Article  Google Scholar 

  5. Balasubramani M, Day BW, Schoen RE, Getzenberg RH . (2006). Altered expression and localization of creatine kinase B, heterogeneous nuclear ribonucleoprotein F, and high mobility group box 1 protein in the nuclear matrix associated with colon cancer. Cancer Res 66: 763–769.

    CAS  Article  Google Scholar 

  6. Ben-Baruch A . (2006). The multifaceted roles of chemokines in malignancy. Cancer Metastasis Rev 25: 357–371.

    CAS  Article  Google Scholar 

  7. Burger JA, Kipps TJ . (2006). CXCR4: a key regulator in the crosstalk between tumor cells and their microenvironment. Blood 107: 1761–1767.

    CAS  Article  Google Scholar 

  8. Chirco R, Liu XW, Jung KK, Kim HR . (2006). Novel functions of TIMPs in cell signaling. Cancer Metastasis Rev 25: 99–113.

    CAS  Article  Google Scholar 

  9. Delacour D, Gouyer V, Zanetta JP, Drobecq H, Leteurtre E, Grard G et al. (2005). Galectin-4 and sulfatides in apical membrane trafficking in enterocyte-like cells. J Cell Biol 169: 491–501.

    CAS  Article  Google Scholar 

  10. Dunlap SM, Celestino J, Wang H, Jiang R, Holland EC, Fuller GN et al. (2007). Insulin-like growth factor binding protein 2 promotes glioma development and progression. Proc Natl Acad Sci USA 104: 11736–11741.

    CAS  Article  Google Scholar 

  11. Dwyer RM, Potter-Beirne SM, Harrington KA, Lowery AJ, Hennessy E, Murphy JM et al. (2007). Monocyte chemotactic protein-1 secreted by primary breast tumors stimulates migration of mesenchymal stem cells. Clin Cancer Res 13: 5020–5027.

    CAS  Article  Google Scholar 

  12. Empereur S, Djelloul S, Di Gioia Y, Bruyneel E, Mareel M, Van Hengel J et al. (1997). Progression of familial adenomatous polyposis (FAP) colonic cells after transfer of the src or polyoma middle T oncogenes: cooperation between src and HGF/Met in invasion. Br J Cancer 75: 241–250.

    CAS  Article  Google Scholar 

  13. Fountoulakis M, Juranville JF, Roder D, Evers S, Berndt P, Langen H . (1998). Reference map of the low molecular mass proteins of Haemophilus influenzae. Electrophoresis 19: 1819–1827.

    CAS  Article  Google Scholar 

  14. Gouyer V, Conti M, Devos P, Zerimech F, Copin MC, Creme E et al. (2005). Tissue inhibitor of metalloproteinase 1 is an independent predictor of prognosis in patients with nonsmall cell lung carcinoma who undergo resection with curative intent. Cancer 103: 1676–1684.

    CAS  Article  Google Scholar 

  15. Grijelmo C, Rodrigue C, Svrcek M, Bruyneel E, Hendrix A, de Wever O et al. (2007). Proinvasive activity of BMP-7 through SMAD4/src-independent and ERK/Rac/JNK-dependent signaling pathways in colon cancer cells. Cell Signal 19: 1722–1732.

    CAS  Article  Google Scholar 

  16. Harbeck M, Schmitt M, Paepke S, Allgayer H, Kates RE . (2007). Tumor-associated proteolytic factors uPA and PAI-1: critical appraisal of their clinical relevance in breast cancer and their integration into decision-support algorithms. Crit Rev Clin Lab Sci 44: 179–201.

    CAS  Article  Google Scholar 

  17. Hu K, Xiong J, Ji K, Sun H, Wang J, Liu H . (2007). Recombined CC chemokine ligand 2 into B16 cells induces production of Th2-dominanted cytokines and inhibits melanoma metastasis. Immunol Lett 113: 19–28.

    CAS  Article  Google Scholar 

  18. Huhtala ML, Pesonen K, Kalkkinen N, Stenman UH . (1982). Purification and characterization of a tumor-associated trypsin inhibitor from the urine of a patient with ovarian cancer. J Biol Chem 257: 13713–13716.

    CAS  PubMed  Google Scholar 

  19. Janicke F, Prechtl A, Thomssen C, Harbeck N, Meisner C, Untch M et al. (2001). Randomized adjuvant chemotherapy trial in high-risk, lymph-node negative breast cancer patients identified by urokinase-type plasminogen activator and plasminogen activator inhibitor type 1. J Natl Cancer Inst 93: 913–920.

    CAS  Article  Google Scholar 

  20. Kazal LA, Spicer DS, Brahinsky RA . (1948). Isolation of a cristalline trypsin inhibitor-anticoagulant protein from pancreas. J Am Chem Soc 70: 3034–3040.

    CAS  Article  Google Scholar 

  21. Kebache S, Ash J, Annis MG, Hagan J, Huber M, Hassard J et al. (2007). Grb10 and active Raf-1 kinase promote Bad-dependent cell survival. J Biol Chem 282: 21873–21883.

    CAS  Article  Google Scholar 

  22. Kikuchi N, Nagata K, Shin M, Mitsushima K, Teraoka H, Yoshida N . (1989). Site-directed mutagenesis of human pancreatic secretory trypsin inhibitor. J Biochem (Tokyo) 106: 1059–1063.

    CAS  Article  Google Scholar 

  23. Klezovitch O, Chevillet J, Mirosevich J, Roberts RL, Matusik RJ, Vasioukhin V . (2006). Hepsin promotes prostate cancer progression and metastasis. Cancer cell 6: 185–195.

    Article  Google Scholar 

  24. Kotelevets L, Noë V, Bruyneel E, Myssiakine E, Chastre E, Mareel M et al. (1998). Inhibition by platelet-activating factor of Src- and hepatocyte growth factor-dependent invasiveness of intestinal and kidney epithelial cells. Phosphatidylinositol 3′-kinase is a critical mediator of tumor invasion. J Biol Chem 273: 14138–14145.

    CAS  Article  Google Scholar 

  25. Kosinski C, Li VSW, Chan ASY, Zhang J, Ho C, Yin Tsui W et al. (2007). Gene expression patterns of human colon tops and basal crypts and BMP antagonists as intestinal stem cell niche factors. Proc Natl Acad Sci USA 104: 15418–15423.

    CAS  Article  Google Scholar 

  26. Le Floch N, Rivat C, De Wever O, Bruyneel E, Mareel M, Dale T et al. (2005). The proinvasive activity of Wnt-2 is mediated through a noncanonical Wnt pathway coupled to GSK-3beta and c-Jun/AP-1 signaling. FASEB J 19: 144–146.

    CAS  Article  Google Scholar 

  27. Lee YC, Pan HW, Penq SY, Lai PL, Kuo WS, Ou YH et al. (2007). Overexpression of tumour-associated trypsin inhibitor (TATI) enhances tumour growth and is associated with portal vein invasion, early recurrence and a stage-independent prognostic factor of hepatocellular carcinoma. Eur J Cancer 43: 736–744.

    CAS  Article  Google Scholar 

  28. Lesuffleur T, Violette S, Vasile-Pandrea I, Dussaulx E, Barbat A, Muleris M et al. (1998). Resistance to high concentrations of methotrexate and 5-fluorouracil of differentiated HT-29 colon-cancer cells is restricted to cells of enterocytic phenotype. Int J Cancer 76: 383–392.

    CAS  Article  Google Scholar 

  29. Loberg RD, Ying C, Craig M, Day LL, Sargent E, Neeley C et al. (2007). Targeting CCL2 with systemic delivery of neutralizing antibodies induces prostate cancer tumor regression in vivo. Cancer Res 67: 9417–9942.

    CAS  Article  Google Scholar 

  30. Louvet C, Coudray AM, Tournigand C, Prévost S, Raymond E, de Gramont A et al. (2000). Synergistic antitumoral activity of combined UFT, folinic acid and oxaliplatin against human colorectal tumor HT29 cell xenografts in athymic nude mice. Anticancer Drugs 11: 579–582.

    CAS  Article  Google Scholar 

  31. Makagiansar IT, Williams S, Mustelin T, Stallcup WB . (2007). Differential phosphorylation of NG2 proteoglycan by ERK and PKCα helps balance cell proliferation and migration. J Cell Biol 178: 155–165.

    CAS  Article  Google Scholar 

  32. McCarthy K, Maguire T, McGreal G, McDermott E, O'Higgins N, Duffy MJ . (1999). High levels of tissue inhibitor of metalloproteinase-1 predict poor outcome in patients with breast cancer. Int J Cancer 84: 44–48.

    CAS  Article  Google Scholar 

  33. Monti P, Leone BE, Marchesi F, Balzano G, Zerbi A, Scaltrini F et al. (2003). The CC chemokine MCP-1/CCL2 in pancreatic cancer progression: regulation of expression and potential mechanisms of antimalignant activity. Cancer Res 63: 7451–7461.

    CAS  PubMed  Google Scholar 

  34. Nguyen QD, De Wever O, Bruyneel E, Hendrix A, Xie WZ, Lombet A et al. (2005). Communtators of PAR-1 signaling in cancer cell invasion reveal an essential role of the Rho-Rho kinase axis and tumor microenvironment. Oncogene 24: 8240–8251.

    CAS  Article  Google Scholar 

  35. Nyberg P, Ylipalosaari M, Sorsa T, Salo T . (2006). Trypsins and their role in carcinoma growth. Exp Cell Res 312: 1219–1228.

    CAS  Article  Google Scholar 

  36. Ohmachi Y, Murata A, Matsuura N, Yasuda T, Yasuda T, Monden M et al. (1993). Specific expression of the pancreatic-secretory-trypsin-inhibitor (PSTI) gene in hepatocellular carcinoma. Int J Cancer 55: 728–734.

    CAS  Article  Google Scholar 

  37. Paju A, Stenman UH . (2006). Biochemistry and clinical role of trypsinogens and pancreatic secretory trypsin inhibitor. Crit Rev Clin Lab Sci 43: 103–142.

    CAS  Article  Google Scholar 

  38. Paju A, Vartiainen J, Haglund C, Itkonen O, von Boguslawski K, Leminen A et al. (2004). Expression of trypsinogen-1, trypsinogen-2, and tumor-associated trypsin inhibitor in ovarian cancer: prognostic study on tissue and serum. Clin Cancer Res 10: 4761–4768.

    CAS  Article  Google Scholar 

  39. Paju A, Hotakainen K, Cao Y, Laurila T, Gadaleanu V, Hemminki A et al. (2007). Increased expression of tumor-associated trypsin inhibitor, TATI, in prostate cancer and in androgen-independent 22Rv1 cells. Eur Urol 52: 1670– press.

    CAS  Article  Google Scholar 

  40. Poole CJ, Earl HM, Hiller L, Dunn JA, Bathers S, Grieve RJ et al. (2006). Epirubicin and cyclophosphamide, methotrexate and fluorouracil as adjuvant therapy for early breast cancer. N Engl J Med 355: 1851–1862.

    CAS  Article  Google Scholar 

  41. Raftopoulou M, Hall A . (2004). Cell migration: Rho GTPases lead the way. Dev Biol 265: 23–32.

    CAS  Article  Google Scholar 

  42. Stenman UH . (2002). Tumor-associated trypsin inhibitor. Clin Chem 48: 1206–1209.

    CAS  PubMed  Google Scholar 

  43. Truant S, Bruyneel E, Gouyer V, De Wever O, Pruvot FR, Mareel M et al. (2003). Requirement of both mucins and proteoglycans in cell-cell dissociation and invasiveness of colon carcinoma HT-29 cells. Int J Cancer 104: 683–694.

    CAS  Article  Google Scholar 

  44. Turpeinen U, Koivunen E, Stenman UH . (1988). Reaction of a tumour-associated trypsin inhibitor with serine proteinases associated with coagulation and tumor invasion. Biochem J 254: 911–914.

    CAS  Article  Google Scholar 

  45. Van der Zee M, Stockhammer O, Von Levetzow C, Nunes da Fonseca R, Roth S . (2006). Sog/Chordin is required for ventral-to-dorsal Dpp/BMP transport and head formation in a short germ insect. Proc Natl Acad Sci USA 103: 16307–16312.

    CAS  Article  Google Scholar 

  46. Xuan JA, Schneider D, Toy P, Lin R, Newton A, Zhu Y et al. (2006). Antibodies neutralizing hepsin protease activity do not impact cell growth but inhibit invasion of prostate and ovarian tumor cells in culture. Cancer Res 66: 3611–3619.

    CAS  Article  Google Scholar 

  47. Yan R, Zu X, Ma J, Liu Z, Adeyanju M, Cao D . (2007). Aldo-keto reductase family 1 B10 gene silencing results in growth inhibition of colorectal cancer cells: implication for cancer intervention. Int J Cancer 121: 2301–2306.

    CAS  Article  Google Scholar 

  48. Yang J, price MA, Neudauer CL, Wilson C, Ferrone S, Xia H et al. (2004). Melanoma chondroitin sulfate proteoglycan enhances FAK and ERK activation by distinct mechanisms. J Cell Biol 165: 881–891.

    CAS  Article  Google Scholar 

  49. Yasuda T, Ogawa M, Murata A, Ohmachi Y, Yasuda T, Mori T et al. (1993). Identification of the IL-6-responsive element in an acute-phase-responsive human pancreatic secretory trypsin inhibitor-encoding gene. Gene 131: 275–280.

    CAS  Article  Google Scholar 

  50. Yasuda T, Yasuda T, Ohmachi Y, Katsuki M, Yokoyama M, Murata A et al. (1998). Identification of novel pancreas-specific regulatory sequences in the promoter region of human pancreatic secretory trypsin inhibitor gene. J Biol Chem 273: 34413–34421.

    CAS  Article  Google Scholar 

Download references


In the memory of Hélène Fontayne-Devaud (TR INSERM) who passed away on 15 April 2007. We thank Dominique Demeyer, Marie-José Dejonghe and Georges Grard for technical assistance, Sabine Quief and Céline Villenet for performing microarrays (Genomic Platform), Marie-Hélène Gevaert and Rosemary Siminsky (Department of Histology, Faculty of Medicine, University of Lille II) and the technicians of the Laboratory of Immunohistochemistry (Center of Biology-Pathology, CHRU-Lille). This work was supported by Cancéropôle Nord-Ouest, La Ligue Contre le Cancer, La Région Nord-Pas de Calais, INSERM and ARC No. 3765.

Author information



Corresponding author

Correspondence to G Huet.

Additional information

Supplementary Information accompanies the paper on the Oncogene website (

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Gouyer, V., Fontaine, D., Dumont, P. et al. Autocrine induction of invasion and metastasis by tumor-associated trypsin inhibitor in human colon cancer cells. Oncogene 27, 4024–4033 (2008).

Download citation


  • mucinous differentiation
  • CXCR4
  • CKB
  • GRB10
  • CSPG4/NG2
  • BMP-7

Further reading