Down-Regulation of Monocyte Chemotactic Protein-3 by Activated β-Catenin

Recent progress in cancer research has underscored the importance of β-catenin, a molecule that plays pivotal roles in cell-to-cell adhesion and in the Wnt/Wg signal-transduction pathway(1). The Wnt/Wg pathway is critical for differentiation and morphogenesis in Drosophila and Xenopus(2), and evidence of aberrant ventralization by ectopic expression of armadillo, the Drosophila homologue ofβ-catenin, has implicated this protein in the determination of cell polarity (3).

Accumulation of β-catenin in the cytoplasm or nucleus as a consequence of mutant APC²or β-catenin (CTNNB1) genes is frequently observed in early stages of colorectal tumorigenesis (4, 5). Moreover deletions of, or mutations within, exon 3 of CTNNB1 have also been identified in tumors of liver, uterus, prostate, skin, and brain (6, 7, 8, 9, 10, 11). Because wild-type APC promotes degradation of intracellular β-catenin through phosphorylation of serine/threonine residues within exon 3, in which β-catenin binds with glycogen synthase kinase 3β (GSK-3β) and Axin/Conductin,mutant forms of APC or β-catenin impair the degradation process. Accumulated β-catenin translocates into the nucleus, in which—in association with Tcf/LEF—it modulates transcription of target genes. Thus far, several transcriptional targets of the β-catenin/Tcf/LEF complex have been identified, including c-myc, cyclin D1, matrilysin,c-Jun, fra-1, uPAR, ZO-1 (12, 13, 14, 15, 16), and NBL-4(17). However, specific molecular targets that are associated with cell differentiation or cell polarity remain to be defined.

A number of chemokines belonging to the C-C subfamily have been identified in the past few years. One of them, MCP-3, is expressed and secreted by monocytes, fibroblasts, platelets, colonic epithelial cells, and some malignant tumor cells (18, 19, 20, 21, 22, 23). MCP-3 activates monocytes, lymphocytes, dendritic cells, natural killer cells, and granulocytes (24). Activation of dendritic cells and type I T cells by MCP-3 can reduce tumorigenicity(25), but this mechanism is not well understood.

To clarify the role(s) of β-catenin in human carcinogenesis, we aimed to identify genes regulated by its activated form (26). We report here that the chemokine MCP-3 is down-regulated by activatedβ-catenin through binding of the nuclear β-catenin complex to a Tcf/LEF-binding sequence (12, 13), which is present in the promoter region of MCP-3. We also describe evidence that down-regulation of MCP-3 may disturb differentiation of colonic cells. Our results bring to light a novel mechanism by which activatedβ-catenin can participate in colorectal tumorigenesis.

We had previously established a mouse fibroblast cell line, l-MT, by introducing into murine L cells a mutantβ-catenin transgene that lacked exon 3 (26). COS-7 cells, human embryonic kidney 293 cell line, and human colon-cancer cell lines HT-29, SW480, SW948, and LoVo were obtained from the American Type Culture Collection (Rockville, MD). All of the cell lines were cultured as monolayers in appropriate media: l-MT, L,and HEK 293 cells in DMEM (Sigma Chemical Co., St. Louis, MO); SW480 and SW948 in Leibovitz’s L-15 and LoVo in Ham’s F12 (Life Technologies, Inc., Grand Island, NY); and HT-29 in RPMI 1640 (Sigma Chemical Co.). All of the media were supplemented with 10% fetal bovine serum (Cansera, Inc., Ontario, Canada) and 1%antibiotic/antimycotic solution (Sigma Chemical Co.), and all of the cells were grown at 37°C in an atmosphere of humidified air containing 5% CO₂.

Total RNA was extracted from L cells, l-MT cells and four human colon-cancer cell lines using TRIZOL Reagent (Life Technologies,Inc.) according to the manufacturer’s protocol. The FDD procedure was performed essentially as described previously (27). PCR products were resuspended in formamide sequencing dye and electrophoresed for 3 h at 1800 V on sequencing gels containing 4% acrylamide (19:1) with 7 m urea. Gel images were analyzed with an FMBIO II Multi-View fluoroimager (TaKaRa, Tokyo,Japan). Bands that showed differential expression between L cells and l-MT cells were excised from the gels, and DNA was extracted by boiling the gel fragments in Tris-EDTA buffer. Each sample was reamplified for 30 cycles with the same primer set used for the FDD procedure. Reamplified products were cloned into pBluescript II SK (−)vector (Stratagene, La Jolla, CA) and sequenced using T3, T7 primers and an ABI PRISM Dye Terminator Cycle Sequencing FS Ready kit(Perkin-Elmer Applied Biosystems Division, Foster City, CA) according to the protocol provided by the supplier.

Cultured cells, replated on chamber slides, were fixed with PBS containing 4% paraformaldehyde for 15 min and were then rendered permeable by incubation for 3 min at 4°C in PBS containing 0.1%Triton X-100. Cells were covered with 2% BSA in PBS for 30 min at room temperature to block nonspecific binding of antibody and then were incubated with a mouse anti-β-catenin antibody (Transduction Laboratories, Lexington, KY). Antibodies were stained with a goat antimouse secondary antibody conjugated to rhodamine (Leinco Technologies, Inc., Ballwin, MO), and viewed with an ECLIPSE E800 microscope (Nikon, Tokyo, Japan).

Expression of the part of APC that corresponds to the 20-amino-acid repeats of its β-catenin-binding domain is able to down-regulateβ-catenin (28). Therefore, we constructed an adenoviral vector containing this domain (Ad-APC) by inserting a 2.5-kb HindIII fragment of APC cDNA into the HindIII site of the pAd-BgIII vector, which contains the cytomegalovirus promoter/enhancer and a bovine growth hormone polyadenylation signal flanked by Ad5 E1 sequences. The recombinant adenoviruses, constructed as described previously(29), were propagated in the HEK293 cell line and purified by two rounds of CsCl density centrifugation. Viral titers were measured by a limiting-dilution bioassay using HEK293 cells. Cell monolayers were infected with the viral solutions and were incubated at 37°C for 1 h, with brief agitation every 15 min. Culture medium was added, and the infected cells were maintained at 37°C for 48 h. Overexpression of APC mRNA was detectable as early as 18 h after infection (data not shown).

Western blotting with mouse anti-β-catenin (Transduction Laboratories) was performed as described elsewhere (30).

A 3-μg aliquot of total RNA from each cell line was reverse-transcribed for single-stranded cDNAs using oligo(dT)₁₅ primer and Superscript II (Life Technologies, Inc., Rockville, MD). Each cDNA mixture was diluted for subsequent PCR amplification by monitoring GAPDH as a quantitative control. The PCR exponential phase was determined on 20–32 cycles to allow comparison among cDNAs developed from identical reactions. As an internal control, the amounts of cDNA were quantified and equalized by amplifying GAPDH. The primer sequences used for amplification were 5′-GACAACAGCCTCAAGATCATCA-3′ and 5′-GGTCCACCACTGACACTGTG-3′ for human GAPDH;5′-CAACTACATGGTTTACATGTTC-3′ and 5′-TGTTCCGAATGTCTGAGGAC-3′ for mouse GAPDH; 5′-TCCAATTCTCATGTTGAAGCC-3′ and 5′-GAGAAAGGACAGGGTATACAAA-3′ for human MCP-3; and 5′-CACTCTCTTTCTCCACCATG-3′ and 5′-GCTAACACAATGTTAAAGTGAC-3′ for mouse MCP-3. All of the reactions involved initial denaturation at 94°C for 2 min followed by 20 cycles (for GAPDH) or by 32 cycles (for MCP-3) at 94°C for 30 s, 57°C for 30 s, and 72°C for 1 min, on a Gene Amp PCR system 9600(Perkin-Elmer). The products were electrophoresed in 3% agarose gels and visualized by fragment Southern-blot analysis followed by the transfer to nylon membranes (Amersham, Cleveland, OH). The membranes were hybridized with ³²P-labeled internal oligonucleotide probes. Each internal oligonucleotide sequence used for Southern-blot analysis as a probe was 5′-CCCAT GGCAAATTCCATGGC-3′ for human GAPDH, 5′-CTCACGGCAAATTC AACGGC-3′ for mouse GAPDH, 5′-GCTACAGAAGGACCACCAGTA-3′ for human MCP-3, and 5′-TTCTGTTCAGGCACATTTCTTC-3′ for mouse MCP-3.

To generate a series of 5′ deletion mutants of the human MCP-3 gene promoter, two fragments were cloned into appropriate enzyme sites of pGL3-Basic Vector (Promega, Madison, WI). Reporter assay was carried out using a Dual-Luciferase Reporter Assay System (Promega, Madison,WI) according to the manufacturer’s protocol. A plasmid vector, pRL-TK(Promega), was cotransfected with each reporter construct into SW480,LoVo, and SW948 cells using FUGENE6 (Boehringer Mannheim, Mannheim,Germany) according to the supplier’s recommendations.

Preparation of nuclear extracts and EMSA were performed essentially as described previously (31). Two pairs of double-stranded,15-nucleotide DNAs were prepared by annealing two oligonucleotide DNAs:5′-ACCAGAATCAAAGCC-3′ and 5′-GGCTTTGATTCTGGT-3′ for the wild-type DNA;and 5′-ACCAGAGCCAAAGCC-3′ and 5′-GGCTTTGGCTCTGGT-3′ for the mutant DNA. The wild-type DNA was end-labeled with polynucleotide kinase in the presence of [γ-³²P]ATP and served as a probe for EMSA. A typical binding reaction contained 10 μg of nuclear extract in 5 μl of extraction buffer [10 mm HEPES (pH 7.9), 400 mm NaCl, 1.5 mm MgCl₂,0.2 mm EGTA, and 20% glycerol], 0.1 ng of the radiolabeled probe, and 100 ng of deoxyinosine-deoxycytidine(dI·dC) in 25 μl of binding buffer (60 mm KCl, 1 mm EDTA, 1 mm DTT, and 10% glycerol). For the competition assays, 50 ng of wild-type or mutant unlabeled DNA was added in the reaction mixture. Samples were incubated for 20 min at room temperature, with or without 0.5 μg of anti-β-catenin antibody and additional incubation for 20 min. Electrophoresis was performed at 4°C on 4% nondenaturing polyacrylamide (29:1) gels with 0.25×Tris-boric/EDTA buffer. The gels were dried and autoradiographed.

The entire coding region of human MCP-3 was cloned into an expression vector, pcDNA 3.1(+) (Invitrogen, Carlsbad, CA), under control of the cytomegalovirus promoter/enhancer. HT-29 colon-cancer cells expressing a high amount of MCP-3 transcript was selected in medium containing 1000 μg/ml geneticin and was subcloned(HT-29-MCP). As a control, HT-29 cells—transfected with the empty vector pcDNA 3.1(+)—were subcloned as well (HT-29-con). All of the cells were grown to 30–50% confluence, either with or without 2 mm sodium butyrate (NaB). After 72 h of incubation,cells were harvested and lysed with lysis buffer. The amount of CEA was analyzed by an ELISA using a commercially available kit (Enzymun-test CEA; Boehringer Mannheim). Measurement of ALP activity was performed as described elsewhere (32).

The data were analyzed using an ANOVA and the Scheffé’s F test.

We previously established a mouse fibroblast cell line, L-MT, in which a mutant form of β-catenin observed in human cancer cells was introduced (26). L-MT cells showed a multilayer growth pattern and displayed growth advantage in low-serum culture media compared with their parent cells, L cells. Expression of β-catenin was up-regulated by the withdrawal of doxycycline in the L-MT cells. Using these cell lines, we aimed to isolate genes regulated byβ-catenin. FDD analysis using l-MT cells revealed a fragment (D15) the intensity of which was decreased in response to the accumulation of β-catenin. DNA sequencing and a subsequent search of the databases for homologies revealed identity of the DNA sequence of D15 to sequences of murine MCP-3. The expression level of murine MCP-3 was inversely correlated with the increase ofβ-catenin in l-MT cells on the withdrawal of doxycycline(Fig. 1,A). That is, when we depleted the culture medium of doxycycline, a significant decrease of MCP-3 expression in l-MT cells was observed within 4 h, and expression became undetectable after 8 h (Fig. 1 B). These results indicated that even a small amount of accumulatedβ-catenin can reduce the expression of MCP-3.

To confirm the inverse correlation between expression of MCP-3 and accumulation of β-catenin, we used a viral vector designed to express the 20-amino-acid-repeat,β-catenin-binding domain of APC (Ad-APC). Infection of these adenoviruses into colon-carcinoma cell line SW480, in which a large amount of β-catenin accumulates in the nucleus and cytoplasm(33), conferred an evident decrease of β-catenin (Fig. 2,A). However, this change was not observed when SW480 cells were infected with adenovirus containing the LacZ gene(Ad-LacZ; Fig. 2,A). It was shown in our previous paper that the reduced expression of β-catenin by Ad-APC was correlated with decreased Tcf/LEF-specific transactivation activity (34). Transfection of the SW480 cells with Ad-APC significantly increased expression of human MCP-3, but no such change occurred in cells transfected with Ad-LacZ (Fig. 2 B).

To identify an element responsible for transcription in the promoter region of hMCP-3, we constructed two reporter-plasmid clones containing different lengths of the region upstream of the luciferase gene (Fig. 3,A). The luciferase activity of reporter-plasmid P1 was significantly lower than that of P2 in the three colon cancer cell lines examined (Fig. 3,B), which indicated the presence of a transcription-suppressing element between −1586 and −924. Because this candidate region contained ATCAAAG, a possible Tcf/LEF-binding motif, we hypothesized that this motif might be responsible for the transcriptional repression. To investigate that hypothesis, we constructed reporter plasmid P1M, in which the candidate Tcf/LEF-binding motif was changed to GCCAAAG (Fig. 3,A). The luciferase assay using these three plasmids revealed that the 1.6-kb fragment containing the mutated motif had lost the ability to suppress transcription of MCP-3; its luciferase activity was equivalent to that of the P2 fragment (Fig. 3 B). These results implied that the putative Tcf/LEF-binding motif is involved in repression of MCP-3 transcription.

To confirm a direct mode of interaction between the β-catenin complex and this promoter element, we prepared a wild-type double-stranded DNA encompassing this sequence and a mutant DNA that involved replacing 2 bp within the motif. Using nuclear extracts from SW480 cells, we performed an EMSA experiment and found that the β-catenin complex bound specifically to this putative element. The specific band was supershifted by the addition of anti-β-catenin antibody (Fig. 4).

To investigate the biological role of MCP-3 in the colon, we transfected hMCP-3 cDNA into colon cancer cell line HT-29,to establish HT-29-MCP3 in which a high level of MCP-3 was constitutively expressed. The HT-29-MCP3 cells showed no apparent difference from parental HT-29 cells in morphology, growth rate, or ability to form colonies (data not shown). However, we found significant differences in regard to CEA production and ALP activity between HT-29-MCP-3 cells and HT-29-con cells transfected with pcDNA 3.1(+) alone. Sodium butyrate (NaB) is known to promote differentiation of colonic epithelial cells (35, 36, 37, 38, 39, 40), and NaB treatment of HT-29 cells increases both the CEA level and ALP activity significantly compared with untreated HT-29 cells. Even in the absence of NaB,HT-29-MCP3 cells revealed 1.7-fold and 1.3-fold excesses of CEA level and ALP activity respectively, compared with HT-29-con cells, whereas NaB treatment of HT-29-MCP3 cells further enhanced production of CEA(Fig. 5,A) and ALP (Fig. 5 B) activity.

The data presented here have demonstrated that MCP-3 is down-regulated by an activated form of β-catenin, and that this decreased expression occurs through direct association of theβ-catenin complex with a putative Tcf/LEF-binding motif present in the MCP-3 promoter.

A number of other mammalian genes including c-myc, cyclin D1, matrilysin (matrix metalloproteinase-7), WISP, c-jun, fra-1, uPAR, ZO-1 (12, 13, 14, 15, 16, 41, 42), and NBL4 (17) are known to be regulated by stabilization and activation of β-catenin. Moreover,several target genes for Wnt signaling have been identified in Xenopus and Drosophila, among them, the nodal-related 3 gene Xnr3 (a member of the transforming growth factor β superfamily); fibronectin; and homeobox genes engrailed, goosecoid, twin, siamois, and ultrabithorax(43, 44, 45, 46, 47, 48). All of them except ZO-1 are transactivated by accumulation of β-catenin. The down-regulation of Tcf-dependent transcription by Groucho, CREB-binding protein (CBP), Sox protein, and NF-κB essential modulator-like kinase (NLK) were also reported, although their association with accumulation ofβ-catenin is unclear at present (49, 50, 51, 52, 53). Down-regulation of genes by Tcf/LEF is thought to reflect one of the following possibilities: (a) Tcf/LEF may bind directly to its binding motif and suppress transcription;(b) because WRM, the homologue of β-catenin in Caenorhabditis elegans, is required to down-regulate the Tcf-like protein POP-1 (54), reduced expression may result from repression of another Tcf-family protein that recognizes a similar binding motif; or (c) the suppression may be a secondary effect of primary targets of Tcf/LEF. These possibilities have not been resolved as yet. However, our data regarding the MCP-3 gene clearly demonstrate that decreased expression by the β-catenin complex is one mechanism by whichβ-catenin regulates downstream genes. Because the Tcf/LEF complex recruits various coactivators or corepressors to modulate transcription, it is conceivable that these associated molecules in combination may determine the function of the complex.

In HT-29 cells, we also found that overexpression of MCP-3 induces CEA and ALP activities, both of which are known to be differentiation markers for the cells (32, 35, 36, 37, 38, 39, 40). We did not detect any morphological differences between MCP-3-transfected HT-29 cells and their parent cells, which may suggest that the effect of MCP-3 alone is not enough to induce detectable morphological changes or that it is involved in a differentiation process not related to microscopic appearance. Hence, repression of MCP-3 may suppress differentiation of the colonic epithelium; this would represent a heretofore-unsuspected mechanism operating in colorectal tumor cells; i.e.,inhibition of differentiation by activated β-catenin. Although the relationship between MCPs and differentiation in colonic cells has not been investigated thoroughly, one group (55) has found that expression of MCP-1 in parenchymal cells was correlated with the histological grade of invasive ductal breast carcinomas. In addition,impaired expression of MCP-1 was involved in cervical tumorigenesis(56). Therefore, the novel role of β-catenin revealed in the experiments documented here has brought a more profound understanding of the mechanisms that underlie colorectal tumorigenesis. Furthermore, controlling MCP-3 expression may represent a means of therapeutic intervention for the treatment of cancer patients in the future.