Galectin-3 is a member of the family of β-galactoside-binding proteins that bind to the carbohydrate portion of cellsurface glycoproteins and glycolipids [1]. Galectin-3 has a chimera-type structure consisting of three different structural domains: a short NH2-terminal domain of 12 amino acids that contains a serine phosphorylation site; a repeated collagen-like sequence that rich in glycine, tyrosine, and proline amino acid residues, which serves as a substrate for matrix metalloproteinases (MMPs); and a COOHterminal carbohydrate recognition domain [1-3]. Galectin-3 is a multifunctional oncogene [1], which regulates cell growth [4], adhesion [5], proliferation [6], angiogenesis [7], and apoptosis [8].

Many studies have shown that galectin-3 regulates cancer cell proliferation. Galectin-3-stimulated cell proliferation of IMR-90 human lung fibroblasts [6]; a decrease of galectin-3 expression in activated T lymphocytes paralleled a downregulation or even a blocking of proliferation [9]; and the introd-uction of galectin-3 cDNA caused human lymphoma Jurkat T cells to grow faster [10]. A recent report provided evidence that downregulation of galectin-3 led to diminished human colon cancer cell proliferation via modulation of the hete-rogeneous nuclear ribonucleoprotein Q (hnRNP Q) level [11].

Overexpression of galectin-3 has been reported in gastric cancer [12]. Positive galectin-3 expression was observed in 84% of gastric cancer cases. In enhanced cells of a cancerous lesion, 48% showed stronger nuclear immunoreactivity than a cytoplasmic one whereas adjacent epithelial cells showed little or weak nuclear immunoreactivity [12]. In addition, decreased galectin-3 expression was found in breast [13], ovary [14], prostate [15], epithelial skin cancer [16], and head-and-neck squamous cell carcinomas [17] than in corresponding normal tissue.

HangAmDan (HAD)-B consists of eight species of Korean medicinal plants and animals (Table 1), and is an upgraded version of HangAmDan (HAD) used traditionally for solid masses, which also shows anti-angiogenic activity [18]. A mixture of these plants has been shown to exert strong anticancer activity against solid tumors, including pancreatic, lung, colorectal, and stomach cancers. Additionally, anti-angiogenesis effects and inhibition of cancer cell proliferation and metastasis have been reported [19]. In particular, case reports observed with HAD have been selected as part of the National Cancer Institute’s Best Case Series Program [20]. HAD-B has shown efficacy in inhibiting migration and proliferation of human umbilical vein endothelial cells and in limiting the formation of capillary tube structures [21]. Furthermore, a safety evaluation of HAD-B has revealed no side-effects in both healthy subjects and cancer patients [22].

Even though a number of studies have reported the functions of galectin-3 in many types of cancer, the mechanisms by which galectin-3 is involved in cell proliferation are not yet fully understood, especially in human colon cancer cells. In the present study we report that γ-aminobutyric acid B receptor 1 (GABABR1) expression is linked to galectin-3 in human colon cancer cell line, and we discuss the effect of galectin-3-independent down-regulation of GABABR1 by treatment with Korean herbal extract HAD-B in human colon cancer cells.

Human colon cancer cell lines, SNU-61, SNU-81, SNU-769B, SNU-C4 and SNU-C5, were obtained from the Korean Cell Line Bank (Seoul, Korea).

[Table 1]

HAD-B was provided from the East-West Cancer Center of Dunsan Oriental Medical Hospital, Daejeon University, Daejeon, Korea (Table 1). The water extract of HAD-B was prepared by extracting HAD-B powder with 10-times (v/w) the amount of distilled water at room temperature for 24 hrs. The extract was centrifuged at 1000×g for 30 mins and was then filtered and lyophilized. The extract powder was dissolved directly in distilled water.

A colorimetric assay using tetrazolium salt, MTT, was used to assess cell proliferation after galectin-3 suppression. MTT assays were performed as described in a previous report [11]. Briefly, equal numbers of cells were incubated in each well in 0.18 ml of culture medium to which 0.02 ml of 10 × 5-FU (Choongwae Pharma Corporation), HAD-B, GABA or PBS (for untreated 100% survival control) had been added. After 4 days of culture, 0.1 mg of MTT was added to each well and incubated at 37℃ for a further 4 hrs. Plates were centrifuged at 450 × g for 5 mins at room temperature, and the medium was removed. Dimethyl sulfoxide (0.15 ml) was added to each well to solubilize the crystals, and plates were immediately read at 540 nm by using a scanning multiwell spectrometer (Bio-Tek Instruments Inc., Winooski, VT). All experiments were performed three times, and the IC50 (μg/ml) values are presented as means ± standard deviations.

Western blot analyses were performed as described in a previous report [11]. Primary antibodies against galectin-3 (Abcam, Cambridge, UK), γ-aminobutyric acid B receptor 1 (GABABR1) (Abcam) and actin (Abcam) (1:1,000) were used.

All procedures were performed at 4℃ unless otherwise specified. Approximately 10⁷ cells in 1 ml of cold 1 × radioimmunoprecipitation assay (RIPA) buffer containing protease inhibitors (Roche Diagnostics) were incubated on ice for 30 mins with occasional mixing. Cell lysates were centrifuged at 12,000 × g for 10 mins, and the supernatant was collected carefully without disturbing the pellet. The supernatant was mixed with primary antibody against either galectin-3 (Abcam) or GABABR1 (Abcam) and was incubated for 2 hrs on a rocking platform. Prepared protein G sepharose beads (GE Health Care Life Sciences) were added and further incubated on ice for 1 hrs on a rocking platform. The mixture was centrifuged at 10,000 g for 30 s, and the supernatant was removed completely. Protein G sepharose beads were washed 5 times with 1 ml of cold 1 × RIPA to minimize the background. Next, 100 μl of 2 × sodium dodecyl sulfate (SDS) sample buffer was added to the bead pellets and heated to 100℃ for 10 mins. After boiling, immunoprecipitates were centrifuged at 10000 × g for 5 mins, and the supernatant was collected for the Western blot analysis.

The intracellular cAMP for human colon cancer cells was determined by using a cAMP Direct Immunoassay Kit (Abcam), as recommended by the manufacturer.

Total RNA was extracted using Trizol reagent (Life Technologies, Inc., Carlsbad, CA), according to the manufacturer’s instructions. Genes expressed in the chemosensitive and chemoresistant groups were analyzed on a high-density oligonucleotide microarray (HG-U133A; Affymetrix, Santa Clara, CA) containing 22,283 transcripts. Target preparation and microarray processing procedures were performed, following the Affymetrix GeneChip Expression Analysis Manual (Affymetrix). Briefly, total RNA extracted was purified with an RNeasy kit (Qiagen). Double-stranded cDNA was synthesized from total RNA (20 μg) with SuperScript II reverse transcriptase (Life Technologies, Inc. Rockville, MD) and a T7-(dT)24 primer (Metabion, Germany). Biotinylated cRNA was synthesized from double-stranded cDNA by using a RNA Transcript Labeling kit (Enzo Life Sciences, Farmingdale, NY), purified, and fragmented. Fragmented cRNA was hybridized to the oligonucleotide microarray, which was washed and stained with streptavidinphycoerythrin. Scanning was performed with an Agilent Microarray Scanner (Agilent Technologies, Santa Clara, CA).

A GeneChip analysis was performed based on the Affymetrix GeneChip Manual (Affymetrix) with Data Mining Tool (DMT) 2.0 and Microarray Database software. All genes represented on the GeneChip were globally normalized and scaled to a signal intensity of 500. Fold changes were calculated by comparing transcripts between the cell lines tested. The DMT 2.0 software employed changed calls (increased or decreased) to analyze the expression of a particular transcript statistically and to determine whether it had been relatively increased, decreased or remained unchanged. After filtration through a "present" call (p< 0.05), a transcript was considered differentially expressed at a fold change of greater than 2.0.

Four genes (ELF3, AXIN2, ENO2 and SACS) were selected for real-time quantitative reverse transcription polymerase chain reaction (qRT-PCR) for validation of the microarray data. Using the SuperScript Pre-amplification System for first strand cDNA synthesis, 5 mg of total RNA was used for creation of singlestranded cDNA (Life Technologies). The cDNA was diluted and quantitatively equalized for PCR amplification. For real-time

[Figure 1] Galectin-3 expression correlated with 5-FU susceptibility in human colon cancer cell lines and gene expression profiling linked to galectin-3. (A) Galectin-3 protein expression correlated with 5-FU susceptibility in three human colon cancer cell lines, SNU-769B, SNU-C4 and SNU-C5. Whole proteomes obtained from the human colon cancer cell lines employed were subjected to SDS-PAGE and were electro-transferred to PVDF membranes for western blot analysis. When galectin-3 expression was higher, human colon cancer cell lines showed more 5-FU susceptibility. (B) Gene expression profiling liked to galectin-3. To satisfy minimum clustering sample size, we added SNU-61, which has almost the same 5-FU susceptibility as SNU-769B, and as shown in the enlarged yellow box, genes linked to galectin-3 expression were selected (a). The expressional profiling was further confirmed by using real-time PCR as shown in panel (b). All genes showing positive and negative expressional correlations with galectin-3 are listed in Tables 2 and 3, respectively.

qRT-PCR, the ABI Prism 7900 sequence detection system (Applied Biosystems) was used. AccuPower GreenStar PCR Master Mix (Bioneer Corporation, Daejeon, Korea) was used for each PCR reaction, and the GAPDH gene was simultaneously run as a control and was used for normalization. Non-template-control wells without cDNA were included as negative controls. Each test sample was run in triplicate. The primer sets for PCR amplification were designed as follows: ELF3-F: 5’-TGAGCTGCTGGAGAAGGATG-3’, ELF3-R: 5’-CCCTTCTTGCAGTCACGAAA-3’, AXIN2-F: 5’-AATCATTCGGCCACTGTTCA-3’, AXIN2-R: 5’-CACAGGCAAACTCATCGCTT-3’, ENO2-F: 5’-CTGATGCTGGAGTTGGATGG-3’, ENO2-R: 5’-CCATTGATCACGTTGAAGGC-3’, SACS-F: 5’-CCATTTGTTGGCATTTTTGG-3’, and SACS-R: 5’-CGCTCATGTTTCAGTGCCTT-3’. Following the standard curve method, the expressed quantities of the examined genes were determined using the standard curves and the CT values and were normalized using the GAPDH expression quantities.

To confirm the correlation between galectin-3 expression and 5-FU susceptibility in human colon cancer cells, we performed Western blot and MTT analyses on three human colon cancer cell lines, SNU-769B, SNU-C4 and SNU-C5. 5-FU susceptibility showed a decreasing tendency that depended on both the transcriptional (Fig. 1A) and the translational (Fig. 1Ba) levels of galectin-3. To cluster the genes showing positively or negatively correlated expression with galectin-3, we employed SNU-61, which had almost the same 5-FU susceptibility as SNU-769B, in a high-throughput gene expression profiling experiment (Fig. 1B & Tables 2, 3). Figure 1Ba shows an example of 19 genes clustered in a galectin-3 expression pattern, which was confirmed by real-time PCR (Fig. 1Bb). The top 50 down- and up-regulated genes in SNU-C5, compared to SNU-769B, are listed in Tables 2 and 3, respectively.

Although the genes listed in Tables 2 and 3 contain γ-aminobutyric acid B receptor 1 (GABABR1), its expression was positi-vely correlated with galectin-3 as previously reported (Fig. 2A) [11]. GABABR1 expression at the translational level was highest in SNU-769B among the three human colon cancer cell lines tested (Fig. 2B). To validate the interaction between galectin-3 and GABABR1, we performed reverse immunoprecipitation: however, galectin-3 did not form a complex with GABABR1 (Fig. 2C).

[Figure 2] Gene and protein expressions of GABABR1 positively linked to galectin-3 expression. (A) GABABR1 in the list of the genes showing positive expressional correlation with galectin-3. (B) Protein expression of GABABR1 in the three human colon cancer cell lines tested. GABABR1 protein expression also showed positive correlation with galectin-3 expression. (C) Reverse immunoprecipitation using anti-galectin and GABABR1 antibody. Results demonstrated that two proteins did not interact to form a complex in SNU-C4 with modest expression of galectin-3 and GABABR1.

To check the effect of HAD-B treatment on the expression level of galectin-3 and GABABR1, we cultured SNU-C4 with modest expression of galectin-3 and GABABR1 in the presence of HADB, and we performed a Western blot analysis. At 96 hrs after treatment with 1 mg/ml HAD-B, expression of GABABR1 was reduced, but galectin-3 did not show any expressional change (Fig. 3A).

Treatment with γ-aminobutyric acid (GABA) in the culture medium promoted proliferation of the human colon cancer cell line SNU-C4 (Fig. 3B). At 48 hrs after treatment with GABA, cell proliferation was increased up to ~50% compared to nonetreated controls, but rate of increase of proliferation was not

[Figure 3] Reduced GABABR1 expression and suppressed cell proliferation of SNU-C4 by treatment with HAD-B. (A) Decreased GABABR1 expression by treatment with HAD-B. At 96 hrs after treatment of with 1 mg/ml HAD-B, protein expression of GABABR1 was decreased in SNU-C4. (B) Suppressed cell proliferation by treatment with HAD-B. GABA treatment recovered the rate of proliferation of SNU-C4 that had been suppressed by HAD-B treatment. (C) Increase in the intracellular cAMP by either GABA or HAD-B treatment. Treatment with GABA or HAD-B increased the basal level of intracellular cAMP.

maintained (Fig. 3B). HAD-B significantly decreased cell proliferation at 48 hrs after treatment compared to the control, but the suppressed proliferation had recovered at 96 hrs (Fig. 3B). Cells co-treated with GABA and HAD-B showed almost the same pattern of proliferation as that of the control (Fig. 3B). Either GABA or HAD-B treatment slightly increased the intracellular cAMP in SNU-C4 compared to that in the nontreated control (Fig. 3C).

[Table 2]

[Table 3]

Colon cancer causes almost a half million deaths every year [23]. In the past 3 decades, 5-fluorouracil (5-FU) chemotherapy and 5-FU-based chemotherapy have been the mainstream in adjuvant treatment of colon cancer [24]; however, partial or complete responses of colon cancer to 5-FU are generally followed by eventual tumor re-growth [25]. Numerous studies have focused on identifying the mechanisms and key molecules involved in natural or acquired 5-FU resistance. Nevertheless, conclusive and consistent results have not been obtained so far. A recent proteome approach identified galectin-3 as a protein affecting 5-FU resistance and the proliferation rate of human colon cancer cells [11]. Our present study confirmed the correlation between galectin-3 expression and 5-FU susceptibility in three human colon cancer cell lines. 5-FU susceptibility of human colon cancer cells was different depending on both the transcriptional and the translational levels of galectin-3 (Figs. 1A and B). Because the identification of genes showing positively or negatively correlated expression with galectin-3 can provide further information on how galectin-3 regulates proliferation of human colon cancer cells, a high-density oligonucleotide microarray was performed. From this transcriptional analysis, we were able to list the genes down- and up-regulated based on the level of galectin-3 expression (Fig. 1B, Tables 2 and 3). Though γ-aminobutyric acid B receptor 1 (GABABR1) was not in the top 50 genes linked to galectin-3 (Tables 2 and 3), interestingly we found that both the transcriptional and the translational levels of GABABR1 were positively correlated with galectin-3 (Fig. 2A and 2B). Even though the biological functions of each individual protein have been well studied, we could not find a report describing the relation between galectin-3 and GABABR1.

GABABRs have been found to play a key role in regulating membrane excitability and synaptic transmission in the brain [26]. GABABRs are G-protein coupled receptors that associate with a subset of G-proteins that trigger cAMP cascades [26]. GABABR subtypes exist; two GABAB-receptor splice variants, GABABR1a and GABABR1b, have been cloned [27], and a new GABABR subtype, GABABR2, does not bind with available GABAB antagonists with measurable potency [28]. GABABR1a, GABABR1b and GABABR2 alone do not activate Kir3-type potassium channels efficiently, but co-expression of these receptors yields a robust coupling to activation of Kir3 channels. GABABR2 and GABABR1a/b proteins immunoprecipitate and localize together at dendritic spines [28]. The heteromeric receptor complexes exhibit a significant increase in agonist- and partial agonist-binding potencies as compared with individual receptors and probably represent the predominant native GABAB receptor [28]. As a previous report also showed that the transcriptional level of GABABR1 was decreased by transfection of galectin-3 small interfering RNA (siRNA) [11], expression of GABABR1 could be regulated by galectin-3. However, reverse immunoprecipitation to validate the interaction between two proteins revealed that galectin-3 did not affect the protein stability of GABABR1 because it formed a complex with GABABR1 (Fig. 2C).

Gamma-aminobutyric acid (GABA) has been reported to affect cancer development. For example, GABA can be a potential tumor suppressor for small airway-derived lung adenocarcinomas [29]. The GABA agonist nembutal has been reported to be a potent inhibitor of primary colon cancer and metastasis [30]. The GABABR agonist baclofen induced G(0)/G(1) phase arrest of human hepatocellular carcinomas (HCCs), which suggested the possibility of developing baclofen as a therapeutic drug for the treatment of HCCs [31]. Furthermore, stimulation of GABABR signaling has been suggested as a novel target for the treatment and the prevention of pancreatic cancer [32]. However, in our present study, treatment with GABA promoted proliferation of the human colon cancer cell line SNU-C4 (Fig. 3B). The Korean herbal extract HAD-B not only decreased GABABR1 expression but also reduced proliferation of human colon cancer cells without any expressional change of galectin-3 (Figs. 3A and 3B). GABABR activation can lead to down-regulation of the intracellular cAMP level in human cancer cells [30,32]. Downregulation of GABABR1 by HAD-B treatment increased the basal level of intracellular cAMP in SNU-C4 (Fig. 3C). However, such an increased cAMP was also observed after GABA treatment (Fig. 3C). The overall findings in the present study were inconsistent with those in previous reports describing the activation of GABABR1 to prevent the progression of a human carcinoma. Nevertheless, our present results showed a link between galectin-3 and GABABR1 in human colon cancer cell proliferation. Galectin-3 regulated GABABR1 expression [11]. Decreased galectin-3 expression reduced not only GABABR1 expression but also the proliferation rate of human colon cancer cells [11]. Even GABA promoted human colon cancer cell proliferation by activating GABABR1 signaling, and the increased proliferation was offset by HAD-B treatment because HAD-B led to galectin-3-independent down-regulation of GABABR1 (Fig. 3).

Our present study confirmed that GABABR1 expression was regulated by galectin-3. Korean herbal extract HAD-B induced galectin-3-independent down-regulation of GABABR1, which resulted in a decreased proliferation of human colon cancer cells. The therapeutic effect of HAD-B for the treatment of human colon cancer needs to be further validated.