SUPPLEMENTARY INFORMATION

SUPPLEMENTARY FIGURE LEGENDS

Supplementary Figure 190K-overexpressing cells havereduced growth and invasion rates.(A)293T and HCT116 cells were stably transfected with 90K cDNA or empty-vector (vec). On the basis of the amounts of 90K secreted in the culture media, which was measured by ELISA assay and subsequently by immunoblot assay, we then selected representative clones for in vitro assays. Myc-tagged90K was examined in cell lysates and serum-free conditioned media (CM) from 90K-overexpressing 293T or HCT116 cells by immunoblot assay using anti-90K antibody (upper) or anti-myc antibody (lower). Glycosylated 90K protein (90 kDa) was detected only in CM, while non-glycosylated 90K protein (74 kDa) in cell lysates (left). Stably 90K secreting (293T/90K, HCT116/90K) cells were analyzed by MTT assay (right). Data are means±SEM (n = 3). * indicates a significant difference between groups(* P < 0.05). (B) Double-culture of Caco2 and HCT116 CRC cells in the upper chamber with 293T/90K or 293T/vector cells in the lower chamber.

Supplementary Figure 2The suppression of the cell invasion rate by lactose was roughly equivalent to that of 90K transfection alone. Comparison of lactose treatment and CM/90K on the suppression of the invasion rate in the HCT116 cells by galectin-containing CM. Values are means±SEM (n=3).* indicates a significant difference between groups(*P<0.05; **P<0.01).

Supplementary Figure 3Interaction of 90K with galectin-1 or galectin-3 was inhibited with increasing doses of lactose.HCT116 cells were treated with sucrose (20 mM) or with increasing doses of lactose, and the cell lysate and cultured media (CM) were prepared, immunoprecipitated with anti- antibody, and immunoblotted with anti-galectin antibody.

Supplementary Figure 4DN-CaMKII, a blockerof CaMKII activity,and S37A-catenin were cotransfected with or without 90K. CaMKII-T286D, an active form of CaMKII, inhibited the elevated TOPFLASH activity by S37A-catenin. Blocking of CaMKII activitywith DN-CaMKII did not restore the TOPFLASH activity reduced by 90K.

Supplementary Figure 5Reduced TOPFLASH reporter activity by 90K in the HCT116 cellswas blocked byMG132 treatment, a blocker ofubiquitin-mediated proteolysis. Also, S37A-catenin degradation by 90K was substantially inhibited when HCT116 cells were treated withMG132.

Supplementary Figure 6Neither KITENIN nor Vangl2 mediates the 90K action on the Wnt/-catenin pathway.90K and β-catenin were cotransfected into 293T cells in the presence ofKITENIN (Vangl1) siRNA orVangl2 siRNA. Treatment with either siRNA did not affect the downregulation of β-catenin-induced TOPFLASH reporteractivity by 90K. * indicates a significant difference between groups(*,P < 0.05).

Supplementary Figure 7 Densitometric analyses of immunoblot for -catenin. The effects of90K on endogenous-catenin level were examined in 293T cellsunder knockdown of CD9 or CD82 via treatment of siRNA following cotransfection of S37A-catenin(left),or under CD9 cotransfection(right). Each -catenin level was normalized with respect to the corresponding actin level. Increased -catenin by S37A-catenin was downregulated after cotransfection of 90K; however, CD9 siRNA completely, whereas CD82 siRNA partially, inhibitedthe downregulation of -catenin by 90K (left). Inversely, cotransfection of90K and CD9 further decreased the-catenin level compared with that of 90K alone (right).Values are mean ± SEM for three independent experiments (n = 3).

Supplementary Figure 8Effects of ganglioside GM3 synthesis blocker on downregulation of β-catenin by 90K.The ganglioside GM3 does not mediate the action of 90K on the Wnt/-catenin pathway. 90K was cotransfected with -catenin (A) orWnt3a (B) or transfected with LiCl (C) into 293T cells, which were pretreated for 24 hours with 20 μM of PDMP,a ganglioside GM3 synthesis blocker, andTOPFLASH reporter activity was examined.

Supplementary Figure 990K showed a suppressive effect on expression of -catenin, cyclinD1, N-cadherin, and number of invading cells in CT-26 cells. (A) CT-26 cells were cotransfected with increasing doses of 90K and S37A-catenin, and TOPFLASH activity was measured. (B) CT-26 cells were cotransfected with increasing doses of 90K and c-fos, andAP-1 reporter activity was measured. (C) Levels of -catenin and cyclinD1 were compared amongCT-26 cell lines that stably expressed the empty-vector (CT-26/vec), 90K cDNA (CT-26/90K), oranti-sense 90K (CT-26/AS-90K) construct. (D) Invasiveness of the CT-26/vec, CT-26/90K, andCT-26/AS-90K cells. Invaded cells were counted in 6 random squares for each filter.

Supplementary Figure 10The negative correlation between the expression level of 90K/CD9/CD82 and advanced stage of human CRC.To assessthe immunohistochemical level of 90K (A), CD9 (B), CD82 (C), -catenin (D), galectin-3 (E), and galectin-1 (F),a score was calculatedby using the intensity of the stainingand the size of the positively stained area.The score indicated theexistence of a negative correlation between the expression level of 90K/CD9/CD82 and stage IV human CRC tissues, and a positive correlation between the expression level of -catenin/galectins and stage IV human CRC tissues.

SupplementalTable 1.PCR primer sequences

Gene / Forward / reverse
90K / gcaaggatcaggccccatca / ggccagcttgtggaagcact
actin / ccggaattcgacaacggctccggcatgtg / cgcgaatccacgatggaggggccggactc
CD9 / accatgccggtcaaaggaggc / accatctcgcggttcctgcgga
cyclinD1 / atggaacaccagctcctgtg / tccatttgcagcagctcctc
DN-APC / cgggatccaggtcaggaacctggaaacgt / gctctagaaacagatgtcacaaggtaaga
DN-Siah1 / cgggatccccctgtaaatatgcgtcttctg / ccgctcgagacacatggaaatagttaca
DN-Siah2 / cgggatcccctattctgcagtgccaggccggg / ccgctcgagctcttgtttctccagcaccag
E-cadherin / ctgctgctgcaggtctcctctt / ccttggccagtgatgctgta
galectin-1 / tcaaacctggggaatgtctc / tcagcctggtcaaaggtgat
galectin-3 / aggagagggaatgatgttgcc / ggtttgccactctcaaaggg
CD82(KAI1) / gaggacaaccagctcattgtg / tcagtacttggggaccttgct
MMP-1 / ttgggctgaaagtgactgg / cctccaatacctgggcctgg
MMP-3 / acaggattgtgaattatacac / gtggccaatttcatgagcagc
N48del-β-catenin / cgggatccggtaaaggcaatcctgaggaa / gctctagacaggtcagtatcaaaccaggc
P53 / atggaggagccgcagtcaga / gtcttggccagttggcaaaaa
-catenin / cgtggacaatggctactcaag / ggatctgcatgccctcatcta
ISG15 / atgggctgggacctgacg / gctccgcccgccaggct

Supplemental Table 2. siRNA target sequences

Gene / siRNA target sequence
90K / 5’-GCCTTTGGTCAAATATTCTTCTGAT -3’
CD9 / 5’-ACAAATGTCTATCAACTTTAA-3’
Gal-1 / 5’-ACCTGAATCTCAAACCTGGAGAGTG-3’
Gal-3 / 5’-TGGGACTCACTCACAGTAACCCTTT-3’
CD82(KAI1) / 5’-TCTTCTACTTCAACATGGGCA-3’
KITENIN / 5’-GGTTGCAGTGGAAGAGGCC-3’
Vangl2 / 5’-TCGTGGTCTCCTACTGGCTCTTCTA-3’

SUPPLEMENTARY METHODS

Cell culture

Human CRC cell lines(HCT116, SW480, HT-29, and Caco2), human embryonic kidney (293T) cells,and mouse colon adenocarcinoma (CT-26) cells were obtained from American Type Culture Collection (Manassas, VA), andwere grown as described.11For the generation of 90K overexpressing human CRC cells, pcDNA6-90K myc-His was transfected into cells by use ofFuGENE6 (Roche) and LipofectAMINE PLUS (Invitrogen) according to the manufacturer's instructions. For the generation of expression or knockdown of 90K in CT-26 cells, pcDNA4-90K-myc or pcDNA4-AS-90K was transfected, respectively. To obtain the secreted galectin, pcDNA6 HA Galectin-1 and -3 were transfected into Caco2 cells. Stable transfectants were selected using 5 μg/ml blasticidin (Invitrogen) for pcDNA6 vector and 100 μg/ml zeocin (Invitrogen) for pcDNA4 vector.For most assays using stable cell lines, we used mixed polyclonal cells to exclude clonal variation.

Plasmids

The coding region of 90KorAS-90K was inserted into the pcDNA6A or pcDNA4expression vector (Invitrogen). TOPFLASH/FOPFLASH, Wnt3a, DN--TrCP, and the wild-typeand S37A -catenin were kindly provided by Dr. Eek-hoon Jho (University of Seoul, South Korea). DN-CaMKII and CaMKII-T286D were kindly provided by Dr. Kunihiro Matsumoto (Kyushu Univ., Japan). AP-1 reporter was provided by Dr. Nacksung Kim (Chonnam National Univ., South Korea).The dominantnegative (DN) constructs weremade by PCR-based methods. Human DN-APC (2543–2843) was constructed by PCRas previously described15 and was subcloned intopcDNA6A (Invitrogen). HumanDN-Siah-1 (Δsiah1;382–940) and DN-Siah2 (Δsiah2; 268-712) were amplified byPCRand clonedinto pIRES-hGFP-1a(Stratagene). ISG15-V5 were amplified by PCR and cloned into pcDNA3.All constructs were confirmed by sequencing.

ELISA and MTT assay

ELISA was used to quantify 90K in cell supernatants. For collecting cell supernatants, empty-vector and 90K-overexpressing cells were cultured without serum in 24-well plates (2×105cells/well).After 2 days, the supernatant was centrifuged and was measured as recommended (Bender MedSystem).

For the MTT (methyl-thiazole tetrazolium bromide) assays, cellswere seeded in 24-well plates (1×105cells/well), and treated with conditioned media from empty-vector- or 90K-overexpressing cells for 48 hours. The assay was performedas described.11

Antibodies, immunoblotting, and immunoprecipitation

Anti-rabbit -cateninand CyclinD1 (Santa Cruz Biotech), and anti-mouse 90K (Bender MedSystem), galectin-1,galectin-3(Abcam), E-cadherin, N-cadherin (BD Pharmingen), CD82, CD9 (Santa Cruz), ubiquitin, anti-actin, anti-HA (Sigma), anti-V5 (Invitrogen), anti-myc (Upstate), and anti-Flag (Sigma) antibodies were used with appropriate secondary antibodies (Amersham). For transient transfection analyses, 293T cells were transfected with various plasmids and were harvested for immunoblot analysis 36 hours after transfection. Cellular proteins were separated, transferred and immunoblotted as described.11 To separate the cytoplasmic and nuclear fraction, NEPER Nuclear and Cytoplasmic Extraction Reagent (Pierce)was used. The blot was reprobed with anti-actin antibody (I-19, Santa Cruz Biotech) to control for loading. To detect membrane bound interaction, DTSSP (Pierce) was used for cross-linking molecules at the cell surface. For immunoprecipitation, proteins were incubated with anti-90K, anti-V5, and anti-HA antibodies and protein A/G-agarose beads (Pierce) and were then analyzed by blotting withspecific antibody.

RNA extraction and RT-PCR

RNA preparation and reverse transcription were performedas described.11 The RT-PCR exponential phase was determined to be 30 cycles to allow for quantitative comparison of the various cDNAs developed from identical reactionson a GeneAmp PCR system(Eppendorf). The PCR primersused are listed in Supplementary Table 1.

RNA interference

For the siRNA interference experiments, we designed an siRNA specific to human 90K, galectin-1,and galectin-3. For the nonspecific siRNA (scramble), we used the All Stars Negative Control siRNA (Qiagen).The siRNA duplexes were prepared andtransfected according to a provided protocol withLipofectamineTM RNAiMAX. Cells were treated for 48 to 72 hours to allow maximum knockdown, followed by RT-PCR and Western blot analysis or use for invasion assays. The siRNA sequencesare listed inSupplementary Table 2.

Cell migration and invasion assay

For the wound healing assays, cellswere seeded in 6-well plates (1×106cells/well), and wounding was performed horizontally with a 1000-µl tip. Migration was recorded immediately after wounding for 72hours at 37°C by using a microscope. Cells were incubated in the absence of serum for 1 hour and were thenincubated with 10% FBS. Wound healingwas quantified by measuring the area migrated across between treatments. Cell invasion was measured by using the Transwell migration apparatus as described.11 Briefly, cultured cells pretreated with siRNA (for 48 hours) or stably overexpressed cells were loaded into the top of a 24-well invasion chamber assay plate (Costar). Conditioned DMEM medium containing 10 μg fibronectin (Calbiochem) was added to the bottom chamber as a chemoattractant. After a 20-hour incubation, the cells were stained. The cells on the top surface of the filters were wiped off with a cotton ball, and the migrated cells on the bottom surface were counted in six random squares of 0.5 mm  0.5 mm for each filter. The results are represented as the mean± SEM of the number of cells per field for at least three independent experiments.

Immunofluorescence and confocal staining for β-catenin or E-cadherin

The pEGFP--cateninplasmid wascotransfected with 90K plasmid, 90K siRNA, or scrambled siRNA into 293T cells by using FuGENE 6. CT-26 cells/vector or /90K or AS-90K, and 293T cells/vector or/90K, and HCT116 cells/vector or/90K were seeded onto an 8-well Lab-Tek Chamber Slide Glass (Nunc) with coated fibronectin and were grown in DMEM supplemented with 10% FBS. Cells were rinsed with phosphate-buffered saline (PBS, pH 7.4) 3 times and fixed with ice-cold 2% buffered paraformaldehyde (pH 7.4) in PBS for 10 min. After washing with PBS and placing with 0.1% Triton X-100 in PBS for 5 min, the cells were incubated with primary antibody, such as anti-β-catenin, or anti-E-cadherin, or anti-N-cadherin for 1 hour at room temperature, and then washed 3 times with PBS buffer. And the cells were incubated with Alexa fluor (Molecular Porbes, A12379) for 30 min at room temperature and were then washed again, and mounted by using Prolong Gold (Molecular Probes, P36939) and were examined with a Laser Scanning Confocal Microscope (Leica).

In vivo tumor growth

Prior approval of the experimental protocol was obtained from the Chonnam National University Medical School Research Institutional Animal Care & Use Committee. Maintenance of animals and all in vivo experiments were performed according to The Guiding Principles in the Care and Use of Animals (DHEW publication, NIH 80-23). CT-26 cell lines(1×106 cells/mouse) that stably expressed the empty-vector (CT-26/vector), that overexpress90K (CT-26/90K), or with knockdownof90K (CT-26/AS-90K) were injected subcutaneously into BALB/c syngeneic mice(n= 6 per group). Tumor size was measured daily with a vernier caliper from the first week to the third week. Tumor volume (V) was calculated by using the values of the largest (A) and the smallest (B) diameter according to the formula: V=0.5AB2. At the third week after injection, the tumor tissues from each experimental group were excised for immunohistochemistry. We also prepared other groups of syngeneic mice that were injected subcutaneously with CT-26/vector, CT-26/90K, or CT-26/AS-90K cells (n= 5 per group). At the sixth week after injection, the presence of metastasis in the liver and lung tissues of each group was evaluated by gross and microscopic examination.

Assessment of immunostaining in CRC specimens

For semiquantification of immunohistochemical90K, CD9, CD82, β-catenin, galectin-3, and galectin-1 expression, a score was attained by multiplying the intensity of the stainingby the size of the area that stained positively. The intensity of cell staining appearing as a dark brown color was graded according to the following scale: 0, no staining; 1+, mild staining; 2+, moderate staining; 3+, marked staining. The area of staining was evaluated on thefollowing scale: 0, <5% of cells stained positive; 1+, 5% to 25% stained positive; 2+, 25% to 50% stained positive; 3+, >50% stained positive. The maximum combined score was 9, and the minimum was 0. Generally, specimens with scores >4 wereconsidered as positive for a candidate protein expression, but this criterion was not used in this study.

Treatment with pharmacologic inhibitors

At 24 hours after transfection or treatment of CM, various inhibitors were added. After 16 hours of incubation with the inhibitors, thecells were lysed for luciferase assay or Western analysis.

Statistical analysis

Experimental differences were tested for statistical significance by using ANOVA and Student’s t test. All statistical tests were two-sided, and P values of less than 0.05 were considered statistically significant.

1