HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 18, 15819-15827, May 3, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/18/15819    most recent M200154200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yu, F. Articles by Kim, H.-R. C. Search for Related Content PubMed PubMed Citation Articles by Yu, F. Articles by Kim, H.-R. C. Social Bookmarking                      What's this?

Galectin-3 Translocates to the Perinuclear Membranes and Inhibits Cytochrome c Release from the Mitochondria

A ROLE FOR SYNEXIN IN GALECTIN-3 TRANSLOCATION^*

Fei Yu, Russell L. Finley Jr.^§, Avraham Raz, and Hyeong-Reh Choi Kim^¶

From the  Department of Pathology and the ^§ Center for Molecular Medicine and Genetics, Wayne State University School of Medicine and Karmanos Cancer Institute, Detroit, Michigan 48201

Received for publication, January 7, 2002, and in revised form, February 11, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Galectin-3 is a multifunctional oncogenic protein found in the nucleus and cytoplasm and also the extracellular milieu. Although recent studies demonstrated an anti-apoptotic activity of galectin-3, neither the functional site nor the mechanism of how galectin-3 regulates apoptosis is known. In this study, we examined the subcellular localization of galectin-3 during apoptosis and investigated its anti-apoptotic actions. We report that galectin-3 translocates to the perinuclear membrane following a variety of apoptotic stimuli. Confocal microscopy and biochemical analysis revealed that galectin-3 is enriched in the mitochondria and prevents mitochondrial damage and cytochrome c release. Using a yeast two-hybrid system, we screened for galectin-3-interacting proteins that regulate galectin-3 localization and anti-apoptotic activity. Synexin, a Ca²⁺- and phospholipid-binding protein, was one of the proteins identified. We confirmed direct interaction between galectin-3 and synexin by glutathione S-transferase pull-down assay in vitro. We showed that galectin-3 failed to translocate to the perinuclear membranes when expression of synexin was down-regulated using an oligodeoxyribonucleotide complementary to the synexin mRNA, suggesting a role for synexin in galectin-3 trafficking. Furthermore, synexin down-regulation abolished anti-apoptotic activity of galectin-3. Taken together, these results suggest that synexin mediates galectin-3 translocation to the perinuclear mitochondrial membranes, where it regulates mitochondrial integrity critical for apoptosis regulation.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Galectin-3 is a 31-kDa member of the -galactoside-binding family of proteins found widely in epithelial and immune cells. Expression of galectin-3 is associated with neoplastic progression and metastatic potential (1-5) in head and neck (6), thyroid (7), gastric (3), and colon (8) cancers, suggesting a role in oncogenesis. Galectin-3 modulates a variety of cellular processes. Extracellular galectin-3 mediates cell migration, cell adhesion, and cell/cell interactions, whereas nuclear galectin-3 is involved in pre-mRNA splicing (9-11). Interestingly, recent studies showed that cytoplasmic, but not nuclear, galectin-3 is associated with tumor progression (12, 13). Yet, the role of cytoplasmic galectin-3 is unknown.

We (15-17) and others (14, 18, 19) have previously shown that galectin-3 inhibits T-cell apoptosis induced by anti-Fas antibody and epithelial cell apoptosis induced by staurosporine, cisplatin, genistein, and anoikis. The anti-apoptotic activity of galectin-3 was also demonstrated in galectin-3-deficient mice. Peritoneal macrophages from galectin-3-deficient mice were more sensitive to apoptotic stimuli than those from control mice (20). The ability of galectin-3 to protect cells against apoptosis induced by agents working through different mechanisms suggests that galectin-3 regulates the common apoptosis commitment step.

During the past decade, explosive progress has been made toward understanding the molecular basis for the regulation of the apoptosis commitment step. Two major apoptotic pathways (intrinsic and extrinsic pathways) have been defined. Intrinsic apoptotic signaling induces cytochrome c release from the mitochondria. Cytosolic cytochrome c initiates the formation of an ~700-kDa complex called the "apoptosome," which consists of cytochrome c, caspase adaptor proteins such as Apaf-1, and caspases (21-23). Apoptosome formation results in caspase activation, a commitment step for apoptosis induction. Extrinsic apoptotic signals are mediated by cell-surface death receptors, including Fas, tumor necrosis factor, and TRAIL receptor families. The death domain of the death receptor initiates the formation of the "death-inducing signaling complex," where caspases are activated (reviewed in Ref. 24). Although a critical role for galectin-3 in apoptosis inhibition is now well documented, neither the functional site nor the molecular mechanism of how galectin-3 regulates apoptosis is understood.

In this study, we investigate the subcellular localization of galectin-3 and its anti-apoptotic actions during intrinsic apoptosis in human breast epithelial cells (BT549). Here, we report that galectin-3 translocates to the perinuclear mitochondrial membranes and inhibits cytochrome c release following a variety of apoptotic stimuli. Whereas caspase activation is drastically down-regulated by galectin-3 overexpression in BT549 cells, exogenous cytochrome c effectively activates caspases in a cell-free system established from galectin-3-overexpressing cells. This suggests that galectin-3 protection of mitochondrial integrity is critical for its ability to down-regulate caspases. We identified synexin as a galectin-3-interacting protein using the yeast two-hybrid system and provide evidence that synexin is involved in galectin-3 translocation to the functional site for its anti-apoptotic actions.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell Culture and Reagents-- The human breast cancer cell line BT549 was obtained from Dr. E. W. Thompson (Vincent T. Lombardi Cancer Research Center, Georgetown University Medical Center, Washington, D. C.). Galectin-3-transfected BT549 cells (BT549Gal-3) were previously established by introducing an expression vector containing human galectin-3 cDNA into parental BT549 cells (15, 16). The neomycin-resistant control vector-transfected BT549 cells are referred to as BT549neo. Cells were cultured using Dulbecco's modified Eagle's medium/nutrient mixture F-12 supplemented with 10% heat-inactivated fetal bovine serum, 100 units/ml penicillin, 100 µg/ml streptomycin, 2 mM L-glutamine, and 0.5 µg/ml Fungizone in a 95% air and 5% CO₂ incubator at 37 °C. All cell culture reagents were purchased from Invitrogen.

DEVDase¹ Activity Assay-- Cells were lysed in cell extract buffer (CEB) (20 mM HEPES (pH 7.5), 10 mM KCl, 1.5 mM MgCl₂, 1 mM EDTA, 1 mM EGTA, and 1 mM dithiothreitol) containing 0.03% Nonidet P-40. Lysates were centrifuged at 15,000 × g for 10 min, and 50 µl of the cytosolic fraction was incubated for 60 min at 37 °C in a total volume of 200 µl of caspase buffer (10 mM HEPES (pH 7.5), 50 mM NaCl, and 2.5 mM dithiothreitol) containing 25 µM acetyl-Asp-Glu-Val-Asp-7-amino-4-methylcoumarin (Bachem, King of Prussia, PA). Using a Spectra Maxi Germini fluorescence plate reader (Molecular Devices, Menlo Park, CA), 7-amino-4-methylcoumarin fluorescence, released by caspase activity, was measured at 460 nm using 360-nm excitation wavelength. Caspase activity was normalized per microgram of protein determined by the BCA protein assay kit (Pierce).

Mitochondrial Staining-- Cells were plated on coverslips in 12-well plates. After 24 h of apoptosis induction, the cells were incubated with medium containing 250 nM MitoTracker Red (Molecular Probes, Inc., Eugene, OR) for 30 min at 37 °C. Cells were washed with PBS and fixed with 3.7% paraformaldehyde in PBS for 15 min at 37 °C. The coverslips were mounted onto glass slides with anti-fade solution (Molecular Probes, Inc.). Fluorescent staining of the mitochondrial membrane was examined with a Nikon Labophot microscope fitted with a digital video camera (Photometrics Ltd., Tucson, AZ) or a Zeiss LSM 310 microscope in the confocal mode.

Cytochrome c Release-- Cells were harvested at 0, 24, and 48 h following treatment with 25 µM cisplatin; resuspended in ice-cold CEB containing 250 mM sucrose and protease inhibitor mixture (Roche Molecular Biochemicals, Mannheim, Germany); and incubated for 1 h at 4 °C. The lysates were then passed through a 261/2-gauge syringe 15 times and then centrifuged at 15,000 × g for 20 min at 4 °C. The resulting supernatant was analyzed by immunoblot analysis using anti-cytochrome c antibody (Zymed Laboratories Inc., South San Francisco, CA). The intensity of the bands was quantified using the Bio-Rad Quantity One program.

Confocal Immunofluorescence Microscopic Analysis-- Cells were cultured on coverslips to 75% confluency. Apoptosis was induced by treatment with 25 µM cisplatin for 24 h or with 0.5 µM staurosporine for 150 min or by growth factor withdrawal for 48 h. Cells were washed with PBS three times, fixed with 3.7% formaldehyde in PBS for 15 min, washed with PBS-S (PBS containing 0.1% saponin), and then incubated with 1% bovine serum albumin in PBS-S for 1 h. After six washes with PBS-S, cells were incubated with rat anti-galectin-3 antibody (American Type Culture Collection, Manassas, VA) or anti-cytochrome c antibody (clone 6H2.B4, BD PharMingen) for 2 h at room temperature. After six washes with PBS-S, the coverslips were incubated with FITC-conjugated secondary antibodies (Sigma) for 1 h. After six washes with PBS-S, the coverslips were mounted upside down with anti-fade solution, sealed, and examined under a Zeiss LSM 310 microscope in the confocal mode.

Immunoblot Analysis-- Cell lysates were prepared using SDS lysis buffer (2% SDS, 125 mM Tris-HCl (pH 6.8), and 20% glycerol). The lysates were boiled for 5 min and then clarified by a 20-min centrifugation at 4 °C. Protein concentration was measured using the BCA protein assay reagent (Pierce). Equal amounts of protein samples in SDS sample buffer (1% SDS, 62.5 mM Tris-HCl (pH 6.8), 10% glycerol, 5% -mercaptoethanol, and 0.05% bromphenol blue) were boiled for 5 min and subjected to reducing SDS-PAGE. After electrophoresis, the proteins were transferred to a nitrocellulose membrane. The blot was blocked with 5% nonfat dry milk in Tris-buffered saline/Tween (100 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.02% NaN₃, and 0.2% Tween 20) for 1 h at room temperature. The membranes were incubated with the appropriate primary antibody in 5% milk in Tris-buffered saline/Tween. After three washes with Tris-buffered saline/Tween, the blot was incubated with the appropriate horseradish peroxidase-conjugated secondary antibody. The antigen was detected using the ECL detection system (Pierce) according to the manufacturer's instruction.

Isolation of Mitochondria-- Mitochondria were isolated as previously described (25). Briefly, BT549Gal-3 and BT549neo cells were homogenized in CEB containing 250 mM sucrose to protect mitochondria by 20 strokes using a type B Dounce homogenizer (Kontes Glass Co., Vineland, NJ). Homogenates were centrifuged at 750 × g for 3 × 10 min at 4 °C to remove debris and nuclei. The supernatant was then centrifuged at 15,000 × g for 20 min; the pellet, which contained mitochondria, was lysed in SDS lysis buffer; and 20 µg of mitochondrial proteins was subjected to immunoblot analysis.

A Cell-free Caspase Activation System-- Cells were lysed in CEB using a Dounce homogenizer as previously described (26). The protein concentration of the lysate was measured using the BCA protein assay kit and adjusted to 4 µg/µl. To activate caspase, cytochrome c (purified from bovine heart; Sigma) was added to cell-free extracts to a final concentration of 50 µg/ml and incubated at 37 °C. DEVDase activity was measured as described above.

Construction of the Bait Plasmid for Yeast Two-hybrid Screening-- The yeast expression vector pEG202 (27), which contains the coding sequences for the LexA DNA-binding domain (amino acids 1-202) and the yeast HIS3 gene, was used to express the bait fusion proteins. The galectin-3 cDNA insert containing the full-length coding sequences was excised from the previously constructed plasmid pcDNA10-galectin-3 (15) and fused in-frame with the LexA DNA-binding domain. The correct orientation and in-frame fusion were confirmed by DNA sequencing. Expression of the fusion proteins was confirmed by immunoblot analysis. The bait plasmid containing the LexA-galectin-3 fusion protein was designated as pLG52.

Yeast Media and Strains-- All yeast media were prepared as described (28). Minimal dropout media contained either 2% glucose (Glc) or 2% galactose (Gal) plus 2% raffinose (Raf). The dropout media lacked uracil, histidine, tryptophan, or leucine and are designated as Ura, His, Trp, or Leu, respectively. Minimal medium containing 0.16 mg/ml X-gal was used to test lacZ reporter gene expression. YPD medium contained yeast extract, peptone, and 2% dextrose. The bait plasmid pLG52 was introduced into Saccharomyces cerevisiae yeast strain RFY206 (MATa his3200 leu2-3lys2201 ura3-52 trp1::hisG) (29) containing the lacZ reporter plasmid pSH18-34 (28) with the yeast URA3 gene by LiOAc-mediated transformation (28). Transformants were selected by growth on GlcUraHis dropout medium. Expression of the fusion protein was confirmed by immunoblot analysis using both anti-LexA and anti-galectin-3 antibodies. To test whether the bait alone would activate the reporter gene LEU2, the bait strain was mated with RFY231 (MAT trp1::hisG his3 ura3-1 leu2::3Lexop-LEU2) (30) containing the TRP1 vector pJG4-5 (27), which was also used to express the cDNA library. The number of diploids grown on Gal/RafUraHisTrpLeu (Leu⁺ colony) and Gal/RafUraHisTrp (colony-forming units) plates was counted. The ratio of Leu⁺ colonies versus total colony-forming units was 7.5 × 10⁷, indicating that the background was low and that the bait plasmid was suitable for yeast two-hybrid screening.

Yeast Two-hybrid Screening-- The human prostate tumor cDNA library cloned into the pJG4-5 plasmid was obtained from OriGene Technology Inc. (Rockville, MD) and maintained in yeast strain RFY231. The bait strain was mated with the RFY231/pJG4-5 cDNA library as previously described (30, 31). Out of 2 × 10⁷ diploid colony-forming units, 244 Leu⁺ colonies were selected. Leu⁺ colonies were printed onto four indicator plates: GlcUraHisTrpLeu, Gal/RafUraHisTrpLeu, Glc/X-galUraHisTrp, and Gal/Raf/X-galUraHisTrp. The colonies that showed galactose-dependent Leu⁺ and LacZ⁺ phenotypes were further analyzed. Plasmids were rescued from the galactose-dependent Leu⁺ and LacZ⁺ colonies by a yeast mini-prep method as previously described (28). PCR amplification was performed using primers BCO1 (5'-CCA GCC TCT TGC TGA GTG GAG ATG-3') and BCO2 (5'-GAC AAG CCG ACA ACC TTG ATT GGA G-3') to amplify the insert. The PCR products were digested with restriction enzymes AluI and HaeIII to avoid sequencing identical clones. The mini-prep DNA was then transformed into Escherichia coli KC8 and purified.

Specificity Test-- The purified prey plasmid DNAs from the galactose-dependent Leu⁺ and LacZ⁺ colonies were introduced back into yeast strain RFY231, and transformants were mated with RFY206/pSH18-34 containing the bait plasmid pLG52 or other randomly chosen baits (LexA fusion bait plasmids pRFHM1 (32), pLex202-hairy (33), p202-DmCdk4,² pKL1,³ and pJG21-1 (27)). Diploids were printed onto four indicator plates as described above. The clones that specifically interacted with pLG52 but not with other baits were sequenced. The sequences were analyzed using the NCBI BLAST search program.

Preparation of Recombinant Galectin-3 Protein-- LB medium containing 10 mM MgCl₂ and 100 µg/ml ampicillin was inoculated with an overnight culture of HMS-174 cells transformed with the plasmid containing the human galectin-3 cDNA insert. When the bacterial cells were grown to an absorbance of 0.5, isopropyl-1-thio--D-galactopyranoside (1 mM) was added. Cells were then incubated for an additional 4 h and harvested by centrifugation at 1250 × g at 4 °C. The pellet was washed with PBS and suspended in 100 mM ice-cold lysis buffer (150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris (pH 8), 0.241 units/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml pepstatin, and 0.2 mM phenylmethylsulfonyl fluoride). The bacterial cells were disrupted by sonication; the lysate was centrifuged at 40,000 × g for 20 min; and the supernatants were passed through an asialofetuin affinity column (22, inner diameter, × 100 mm). The column was made by linking the asialofetuin to Affi-Gel 15 (Bio-Rad) according to the manufacturer's protocol and was well equilibrated in phosphate buffer (10 mM phosphate, 1 mM MgSO₄, 0.2 mM phenylmethylsulfonyl fluoride, and 0.2% NaN₃ (pH 7.2)). The column was washed with 3-5 column volumes of phosphate buffer, and the bound protein was eluted with 0.2 M lactose. Eluted fractions were quantitated using the BCA protein assay reagent and analyzed by SDS-PAGE and immunoblot analysis using anti-galectin-3 antibody.

Preparation of the GST-Synexin Fusion Protein and in Vitro Binding Assay-- The expression plasmid pGEX-KG-synexin for the GST-synexin fusion protein was obtained from Dr. Carl E. Creutz (University of Virginia). The GST expression plasmid pGST-4T3 was obtained from Dr. Brooks (Wayne State University). The GST and GST-synexin fusion proteins were prepared as previously described (34). Briefly, expression of the GST and GST-synexin fusion proteins was induced in E. coli XL-1Blue cells by 100 µM isopropyl-1-thio--D-galactopyranoside for 5 h. The cells were collected by centrifugation (5000 × g, 10 min), resuspended in 50 ml of PBST1 (1× PBS with 1% Triton X-100), and lysed by sonication. The lysate was centrifuged at 10000 × g for 10 min. The supernatant was incubated with 1 ml (50% slurry) of reduced GSH-Sepharose (Amersham Biosciences). After washing the beads six times with PBS, the binding proteins were eluted three times with 1 ml of 10 mM reduced glutathione. The eluted proteins were dialyzed against PBS overnight at 4 °C using Slide-A-Lyzer 10K dialysis cassette (Pierce). In vitro binding of galectin-3 to synexin was examined by GST pull-down assay as previously described (35). Briefly, GSH-Sepharose beads were pretreated with bacterial lysates. 1 µg of GST or GST-synexin proteins in 500 µl of PBST1 was absorbed to 50 µl of the pretreated beads (50% slurry) for 2 h, and then the beads were washed three times with PBST1. Then, 1 µg of recombinant human galectin-3 proteins in 500 µl of PBST05 (1× PBS with 0.05% Triton X-100) was incubated with the GST or GST-synexin beads in the absence or presence of 20 µg of bovine serum albumin at room temperature for 2 h. In a control experiment, 1 µg of galectin-3 protein in 500 µl of PBST05 was heat-denatured by boiling for 5 min, followed by quenching on ice. After washing the beads six times with PBST05, the binding proteins were eluted with 25 µl of SDS sample buffer and subjected to SDS-PAGE and immunoblot analysis.

Preparation of Antisense and Scrambled Oligonucleotides-- The phosphorothioate oligonucleotide (5'-GGA TAG CCT GGG TAT GAC ATT C-3'), complementary to the human synexin mRNA sequences surrounding the ATG translation start site, and the control oligonucleotides containing scrambled nucleotide sequences (5'-CTT ACA GTA TGG GTC CGA TAG G-3') were synthesized and purified by high performance liquid chromatography at Integrated DNA Technologies, Inc. (Coralville, IA). To detect cells into which the oligomers were introduced, the antisense oligonucleotides were labeled with tetramethylrhodamine at the 3'-end (Integrated DNA Technologies, Inc.).

Transient Transfection of Oligonucleotides into BT549Gal-3 Cells-- Cells were transfected with oligonucleotides using Effectene (QIAGEN Inc., Chatsworth, CA) as instructed by the manufacturer. Briefly, 2 µg of antisense or scrambled oligonucleotides was mixed with 120 µl of EC buffer and 16 µl of enhancer by vortexing for 1 s, followed by incubation for 4 min at room temperature. The DNA was vortexed with 20 µl of Effectene for 1 min, followed by incubation for 10 min. The DNA/Effectene solution was mixed with 800 µl of growth medium and overlaid onto cells grown in six-well plates to 70% confluency. In control experiments, cells were treated with the Effectene mixture without oligonucleotides or with antisense or scrambled oligonucleotides without Effectene. After 24 h, the cells were harvested and lysed in SDS lysis buffer for immunoblot analysis or in caspase lysis buffer for caspase activity assay. For transfection with tetramethylrhodamine-labeled antisense oligonucleotides, cells grown on a coverslip in a 12-well plate were treated with the DNA/Effectene mixture (1 µg of oligonucleotide, 60 µl of EC buffer, 8 µl of enhancer, 10 µl of Effectene, and 400 µl of medium). After 6 h of transfection, the medium was replaced with fresh growth medium; and 24 h later, the cells were treated with 0 and 0.5 µM staurosporine for 150 min and then stained with anti-galectin-3 antibody/FITC-conjugated secondary antibody.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Galectin-3 Protects Mitochondrial Integrity-- Mitochondrial events critical for apoptosis include the disruption of electron transport, loss of mitochondrial transmembrane potential, and release of cytochrome c (23, 36), resulting in caspase activation. To examine whether galectin-3 protects mitochondrial integrity, we stained BT549neo and BT549Gal-3 cells with MitoTracker Red, which selectively stains mitochondria and serves as a marker for the mitochondrial membrane potential (37). Overexpression of galectin-3 in BT549Gal-3 cells was confirmed by immunoblot analysis (Fig. 1A). Thirty-six hours of treatment with 25 µM cisplatin resulted in the loss of mitochondrial structure in BT549neo cells (Fig. 1, B and C). In contrast, the mitochondria in cisplatin-treated BT549Gal-3 cells retained the fibrillar fluorescence pattern, as observed in the untreated cells (Fig. 1, D and E), suggesting that galectin-3 overexpression protects cells against the loss of mitochondrial transmembrane potential. As predicted from the loss of mitochondrial integrity, the immunoblot analysis of cytosolic cytochrome c showed that the level of cytochrome c released from the mitochondria was elevated in BT549neo cells compared with BT549Gal-3 cells following cisplatin treatment (Fig. 1F). After 48 h of treatment, a >10-fold increase in cytosolic cytochrome c was detected in BT549neo cells, whereas only an ~1.6-fold increase was observed in BT549Gal-3 cells. To exclude the possibility that cytochrome c was released to the cytosolic fractions during the preparation of cellular homogenates, we performed immunostaining of cytochrome c in control and apoptotic cells. The majority of BT549neo cells exhibited diffuse cytochrome c staining following treatment with cisplatin (Fig. 2B) or staurosporine (Fig. 2D) or after growth factor withdrawal (Fig. 2C). In contrast, cytochrome c staining in BT549Gal-3 cells remained punctate following the same treatment (Fig. 2, F-H). Confocal immunofluorescence microscopic analysis of cells co-stained with anti-cytochrome c antibody/FITC-conjugated secondary antibody and with MitoTracker Red confirmed that cytochrome c remained in the mitochondria in BT549Gal-3 cells following apoptotic stimuli, as shown by yellow staining (Fig. 2, N-P). These results show that galectin-3 overexpression protects cells from losing their mitochondrial membrane potential and prevents cytochrome c release to the cytosol following a variety of apoptotic stimuli.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Galectin-3 expression in human breast epithelial cells prevents mitochondrial damage and cytochrome c release. A, immunoblot analysis of galectin-3 was performed using 20 µg of cell lysates from parental BT549, BT549neo, and BT549Gal-3 cells. The same blot was reprobed with anti--actin antibody to confirm the equal loading of proteins in each lane. B-E, BT549neo (B and C) and BT549Gal-3 (D and E) cells were cultured on coverslips and treated with 0 (B and D) or 25 (C and E) µM cisplatin. After 36 h, the cells were stained with a fluorescent probe for the mitochondrial membrane potential (MitoTracker Red). The fluorescence study was carried out with a Nikon Labophot microscope fitted with a digital video camera. F, immunoblot analysis of cytosolic cytochrome c (cyt c) was performed using cytosolic proteins prepared from BT549Gal-3 and BT549neo cells treated with or without 25 µM cisplatin for 24 or 48 h. To confirm the equal loading of proteins in each lane, the same blot was reprobed with anti--actin antibody.

View larger version (145K): [in this window] [in a new window]   Fig. 2.   Galectin-3 expression in human breast epithelial cells inhibits cytochrome c release from mitochondria in response to apoptotic stimuli. BT549neo (A-D and I-L) and BT549Gal-3 (E-H and M-P) cells were cultured on coverslips with no treatment (A, E, I, and M), with 25 µM cisplatin for 24 h (B, F, J, and N), with serum-free medium for 48 h (C, G, K, and O), or with 0.5 µM staurosporine for 150 min (D, H, L, and P). The mitochondria and cytochrome c were stained as described under "Materials and Methods." A-H show cytochrome c staining only; I-P are composite images of cytochrome c (green) and mitochondria (red), with the yellow color indicating co-localization.

Galectin-3 Inhibition of Cytochrome c Release Is Critical for Its Inhibition of Caspase Activation-- We previously showed that galectin-3 overexpression results in inhibition of poly(ADP-ribose) polymerase cleavage following apoptotic stimuli (15, 16), suggesting that galectin-3 down-regulates caspase activation. Because effector caspases (such as caspase-3 and -7) cleave poly(ADP-ribose) polymerase at the DEVD²¹⁶G site, we measured cisplatin- and staurosporine-induced DEVDase activity in BT549neo and BT549Gal-3 cells using the fluorogenic substrate acetyl-DEVD-7-amino-4-methylcoumarin. DEVDase (caspase-3-like) activity increased ~5-fold at 48 h following 25 µM cisplatin treatment in BT549neo cells, whereas it increased only ~2-fold following the same treatment in BT549Gal-3 (Fig. 3A). Similarly, DEVDase activity increased ~5-fold at 2.5 h following 0.5 µM staurosporine treatment in BT549neo cells, whereas no significant increase was detected in BT549Gal-3 cells (Fig. 3B). These results show that the caspase-3 (effector caspase)-like activity necessary for apoptosis execution is significantly inhibited by galectin-3 overexpression. We then tested whether galectin-3 inhibition of caspase activity results from changes in the apoptotic machinery necessary for caspase activation or from inhibition of cytochrome c release. To this end, we established a cell-free caspase activation system as previously described (26). Caspases in extracts from human breast epithelial cells were effectively activated by adding cytochrome c at 50 µg/ml, and the activation kinetics were comparable between BT549neo and BT549-Gal-3 cell-free extracts (Fig. 3C). Surprisingly, cell-free extracts from BT549Gal-3 cells contained caspases that could be activated at even higher levels in the presence of cytochrome c compared with the BT549neo cell extracts. These results indicate that BT549Gal-3 cells contain all the necessary components to activate caspases and suggest that galectin-3 down-regulation of caspases largely results from its ability to inhibit cytochrome c release following apoptotic stimuli.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Galectin-3 expression in human breast epithelial cells down-regulates DEVDase activity. A and B, BT549neo and BT549Gal-3 cells were treated with 25 µM cisplatin for 48 h (A) or with staurosporine for 150 min (B). DEVDase activities were measured and normalized per microgram of protein. Three independent experiments were performed, and the error bars represent the mean ± S.D. of triplicates. The level of DEVDase activity in untreated cells was arbitrarily given as 1. C, cell-free caspase activation systems were established from BT549neo and BT549Gal-3 cells as described under "Materials and Methods." Cell-free extracts of BT549neo and BT549Gal-3 cells were incubated with or without exogenous cytochrome c (cyt c; 50 µg/ml) for the indicated time periods, and DEVDase activities were measured. RFU, relative fluorescence units.

Galectin-3 Is Redistributed onto the Intracellular Membranes following Apoptotic Stimuli-- To determine the subcellular location where galectin-3 exerts its anti-apoptotic effect, the galectin-3 protein in the control and apoptotic cells was stained with anti-galectin-3 monoclonal antibody and FITC-conjugated secondary antibody. Galectin-3 staining was evenly detected in the nucleus and cytoplasm in the control BT549Gal-3 cells (Fig. 4A). In contrast, galectin-3 staining displayed a compact and punctate extranuclear membrane staining in cells treated with cisplatin (Fig. 4B), cultured in serum-free medium (Fig. 4C), or treated with staurosporine (Fig. 4D). This suggests that following apoptotic stimuli, galectin-3 translocates to the intracellular membrane, possibly to the mitochondria, where it prevents mitochondrial dysfunction and inhibits caspase activation. To examine whether galectin-3 translocates to the mitochondria, cells were co-stained with anti-galectin-3 monoclonal antibody and MitoTracker Red. The galectin-3 staining patterns (Fig. 4, B-D) strikingly resembled the mitochondrial staining patterns (Fig. 4, F-H) following apoptotic stimuli, but not those in the control cells (Fig. 4, A and E). Confocal microscopic analysis from the same plane of focus revealed that galectin-3 and mitochondria indeed co-localized following apoptotic stimuli, as shown by yellow staining (Fig. 4, J-L). To further confirm galectin-3 translocation to the mitochondria, mitochondria were isolated in the presence of 250 mM sucrose as previously described (25). Immunoblot analysis confirmed a significant increase in the level of galectin-3 protein in the mitochondrial fraction following apoptotic stimuli (Fig. 4M).

View larger version (60K): [in this window] [in a new window]   Fig. 4.   Galectin-3 is redistributed onto the intracellular membranes following apoptotic stimuli. A-L, BT549Gal-3 cells treated with no apoptotic stimulus (A, E, and I), with 25 µM cisplatin for 24 h (B, F, and J), with serum-free medium for 48 h (C, G, and K), or with 0.5 µM staurosporine for 150 min (D, H, and L) were co-stained with MitoTracker Red and anti-galectin-3 antibody/FITC-conjugated secondary antibody. Galectin-3 (green staining; A-D) and mitochondria (red staining; E-H) are shown. Co-localization of galectin-3 and mitochondria (yellow staining; I-L) is shown by composite images of cells (indicated by arrows) at a higher magnification. M, mitochondria were isolated from BT549Gal-3 cells treated with no apoptotic stimulus (control (Ctr)), with serum-free medium for 48 h (SF), with 25 µM cisplatin for 24 h (Cisp), or with 0.5 µM staurosporine for 150 min (STS). The levels of galectin-3 proteins in the mitochondria were detected by immunoblot analysis (upper panel). To confirm the equal loading of the mitochondrial proteins in each lane, the same blot was probed with anti-cytochrome c (cyt c) oxidase antibody (Molecular Probes, Inc.) (lower panel).

Galectin-3 Interacts with Synexin-- The galectin-3 protein lacks signal sequences for its subcellular localization, suggesting that galectin-3 translocation may occur through its interaction with other proteins that direct protein trafficking. To search for proteins that interact with full-length galectin-3, we screened a human prostate tumor cDNA library using the LexA yeast two-hybrid screening methods as described (27, 28, 30). Out of 2 × 10⁷ diploid colony-forming units screened, 17 positive colonies were detected. cDNA plasmids were isolated from the positive colonies and amplified in E. coli for further analyses. The purified plasmids were introduced back into yeast cells and tested for specific interactions by performing two-hybrid interaction mating assays with the LexA-galectin-3 bait or with five other randomly chosen baits. Eleven of the 17 candidates interacted with galectin-3, but not with the five randomly chosen baits. DNA sequencing analysis revealed that 2 of the 11 prey plasmids encoded for the full-length synexin protein. Yeast two-hybrid interaction mating assays showed specific interactions between galectin-3 and synexin (Fig. 5A). The remaining nine prey plasmids represented two novel proteins that will be described elsewhere.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Galectin-3 interacts with synexin. A, yeast two-hybrid interaction mating assay. Yeast RFY206/pSH18-34 containing the LexA-galectin-3 bait (Gal-3) or five randomly chosen baits (pRFHM1 (32) (lane 1), plex202-hairy (lane 2), p202-DmCdk4 (lane 3), pKL1 (lane 4), or pJG21-1 (lane 5)) was mated with RFY231 containing the synexin prey. Diploids were replicated onto UraHisTrpLeu medium (leu; upper panels) or UraHisTrp medium containing X-gal (X-Gal; lower panels). The left panels contained galactose, and the right panels contained glucose. The specific interaction between galectin-3 and synexin (Syn) was detected by galactose-dependent growth on the Leu plate and by galactose-dependent -galactosidase expression as indicated by the blue colony on the X-gal plate. B, GST pull-down assay. 1 µg of GST (first lane) or GST-synexin fusion proteins (second through fourth lanes) bound to GSH-Sepharose beads were incubated with 1 µg of recombinant human galectin-3 proteins (Gal-3) in the absence (first and second lanes) or presence (third lane) of 20 µg of bovine serum albumin (BSA) or with 1 µg of heat-denatured galectin-3 proteins (fourth lane). Galectin-3 proteins eluted from the beads were detected by immunoblot analysis using anti-galectin-3 antibody.

To confirm the direct interaction between galectin-3 and synexin, we performed an in vitro binding assay. GST and GST-synexin fusion proteins were purified from the bacterial expression system, and direct interactions between synexin and recombinant galectin-3 proteins were tested by GST pull-down assays as described under "Materials and Methods." As shown in Fig. 5B, galectin-3 bound directly to GST-synexin, but not to GST. Synexin effectively interacted with galectin-3 in the presence of a large excess of bovine serum albumin, whereas it failed to bind to heat-denatured galectin-3 proteins, indicating that galectin-3 binding to synexin is specific.

Synexin Is Critical for Galectin-3 Translocation to the Perinuclear Membrane and Apoptosis Regulation-- Synexin (annexin VII) is a member of the annexin Ca²⁺- and phospholipid-binding family of proteins (38). Synexin is thought to act as a Ca²⁺ channel and as a Ca²⁺-activated GTPase, thus regulating Ca²⁺/GTPase-dependent secretory events (39, 40). To examine the significance of synexin for galectin-3 trafficking and apoptosis regulation, we down-regulated synexin expression in BT549Gal-3 cells using antisense oligonucleotides. Transfection of the oligonucleotides complementary to the synexin mRNA significantly down-regulated synexin expression, whereas control oligonucleotides (scrambled sequences) had no effect (Fig. 6). When synexin expression was down-regulated, intracellular galectin-3 levels also decreased, whereas extracellular galectin-3 levels increased (~5-fold), suggesting that synexin is involved in galectin-3 trafficking.

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Synexin expression is critical for galectin-3 trafficking. BT549Gal-3 cells were treated with Effectene without oligonucleotides (Control) or with scrambled oligonucleotides (SC oligo) or antisense oligonucleotides complementary to the synexin mRNA (AS oligo) in the absence (EF) or presence of (+EF) Effectene for 24 h. The synexin and galectin-3 levels in total cell lysates were detected by immunoblot analysis using anti-synexin and anti-galectin-3 antibodies, respectively. To confirm the equal loading of proteins in each lane, the same blot was probed with anti--actin antibody. Extracellular galectin-3 levels were detected by immunoblot analysis using conditioned medium (CM).

We then examined whether synexin expression is critical for subcellular redistribution of galectin-3 following apoptotic stimuli. To detect cells into which oligonucleotides were introduced, cells were transfected with the antisense oligonucleotides labeled with tetramethylrhodamine at the 3'-end. Immunostaining with anti-galectin-3 antibody/FITC-conjugated secondary antibody showed that galectin-3 staining was evenly detected in the nucleus and cytoplasm in antisense oligonucleotide-transfected BT549Gal-3 cells (Fig. 7, A and C) as in control BT549Gal-3 cells (Fig. 4A). Cells with no or small amounts of the antisense oligonucleotide contained high levels of intracellular galectin-3 proteins (Fig. 7, A and C, single-tailed arrows), whereas cells with high levels of the antisense oligomers retained lower levels of intracellular galectin-3 (double-tailed arrows). These results are consistent with those obtained by immunoblot analysis showing that synexin down-regulation resulted in down-regulation of intracellular galectin-3 levels (Fig. 6). Following apoptotic stimuli, galectin-3 was localized to the perinuclear membranes in BT549Gal-3 cells (Fig. 4D), whereas it remained evenly distributed in cells transfected with antisense oligonucleotides (Fig. 7B). Redistribution of galectin-3 proteins was consistently shown in BT549 cells into which antisense oligonucleotides were introduced at low efficiency (Fig. 7, B and D, insets). These results indicate that synexin expression is critical for intracellular galectin-3 expression and redistribution during apoptosis. We then tested whether disruption of galectin-3 translocation by down-regulation of synexin has effects on the anti-apoptotic activity of galectin-3. We measured staurosporine-induced DEVDase activity in BT549Gal-3 cells with or without synexin down-regulation (Fig. 7E). The basal levels of DEVDase activity in BT549Gal-3 cells were not significantly altered by synexin down-regulation using antisense oligonucleotides. However, DEVDase activity increased >3-fold following staurosporine treatment when synexin expression was down-regulated. This increase was similar to the level of DEVDase activation in BT549neo cells following the same treatment (Fig. 3B), showing that synexin down-regulation abolishes the ability of galectin-3 to down-regulate caspase activity.

View larger version (25K): [in this window] [in a new window]   Fig. 7.   Synexin expression is critical for galectin-3 translocation and apoptosis regulation in human breast epithelial cells. A-D, BT549Gal-3 cells transfected with tetramethylrhodamine-labeled antisense synexin oligonucleotides were treated with 0 (A and C) or 0.5 µM staurosporine for 150 min (B and D), stained with anti-galectin-3 antibody/FITC-conjugated secondary antibody, and examined under a confocal microscope. Galectin-3 (green staining) and transfected tetramethylrhodamine-labeled oligonucleotides (red staining) are shown. In A and C, single-tailed arrows indicate cells with low levels of tetramethylrhodamine-labeled oligonucleotides, and double-tailed arrows indicate cells with high levels of oligonucleotides. The insets in B and D represent a staurosporine-treated cell into which tetramethylrhodamine-labeled oligonucleotides were introduced at a low level. E, BT549Gal-3 cells were treated with Effectene without oligonucleotides (Control), with scrambled oligonucleotides without Effectene (SC), with antisense oligonucleotides without Effectene (AS), with scrambled oligonucleotides with Effectene (SC+EF), or with antisense oligonucleotides with Effectene (AS+EF). After treatment with 0 or 0.5 µM staurosporine for 150 min, DEVDase activities were measured and normalized per microgram of protein and are presented as relative fluorescence units (RFU). Three independent experiments were performed, and the error bars represent the mean ± S.D. of triplicates.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Galectins are a family of evolutionarily conserved animal lectins. During the past decade, efforts have been made to dissect the multiple functions of galectins. Recent studies including ours revealed that some members of the galectin family of proteins are novel regulators of apoptosis (14-16, 41-43). The galectin-7 gene is an early transcriptional target of the tumor suppressor gene product p53 following genotoxic stresses such as UV irradiation, and galectin-7 overexpression enhances keratinocyte apoptosis (44, 45). The extracellular galectin-1 and -9 proteins induce apoptosis in thymocytes or activated T-cells (41, 42). In contrast, our studies (15, 16) and others (14) indicate that intracellular, but not extracellular, galectin-3 inhibits apoptosis. When the recombinant galectin-3 proteins were supplemented in the medium at a concentration of up to 2 µg/ml, they failed to protect BT549 cells against apoptosis. Similarly, the conditioned medium containing galectin-3 secreted by galectin-3-overexpressing cells (BT549Gal-3) failed to down-regulate apoptotic events in BT549 cells, suggesting that extracellular galectin-3 lacks anti-apoptotic activity (data not shown). This study consistently indicates that cytoplasmic galectin-3 regulates epithelial cell apoptosis. In conjunction with recent clinical studies showing that cytoplasmic galectin-3 correlates with tumor progression (12, 13), this study may provide an explanation for the oncogenic activity of cytoplasmic galectin-3 in human carcinoma cells.

Following apoptotic stimuli, galectin-3 is enriched in the intracellular membrane including the mitochondria, a key regulation site for apoptosis induced by a variety of stimuli. Galectin-3, a non-Bcl-2 family member, effectively protects mitochondrial integrity and down-regulates the caspase cascade following intrinsic apoptotic signals. Many Bcl-2 family proteins reside in the mitochondrial outer membrane, endoplasmic reticulum, and nuclear envelope through their C-terminal membrane anchor domains (reviewed in Ref. 46). The Bcl-2 family proteins in the mitochondrial membrane are thought to interact with the permeability transition pore complex and to regulate the opening of the conductance channel. Unlike Bcl-2 family members, galectin-3 does not have a membrane anchor domain. Structurally, galectin-3 is composed of two distinct domains: an N-terminal domain containing proline- and glycine-rich sequences and a globular C-terminal domain containing the carbohydrate recognition site (5). Galectin-3 contains four amino acid residues (NWGR) that are conserved in the Bcl-2 homology domain 1 (BH1) of the Bcl-2 family. This motif is critical for Bcl-2 anti-apoptotic activity and its interaction with other Bcl-2 family members (47). Similar to the Bcl-2 protein, substitution of Gly¹⁸² with Ala in the NWGR motif of galectin-3 abrogates its anti-apoptotic function (15, 16). The galectin-3 protection of mitochondrial integrity may result from its ability to interact with Bcl-2 family members, as previously suggested (14), or galectin-3 may directly interact with the mitochondrial permeability transition pore complex and regulate its opening. This study shows that, although its molecular action is still unclear, galectin-3 exerts its anti-apoptotic activity at the perinuclear mitochondrial membranes. Caspases in cell-free extracts prepared from BT549Gal-3 cells can be effectively activated by addition of cytochrome c. This suggests that once significant amounts of cytochrome c are released from the mitochondria, galectin-3 fails to inhibits apoptotic events. Thus, galectin-3 inhibition of cytochrome c release from the mitochondria is critical for apoptosis inhibition.

Synexin (annexin VII), a 51-kDa member of the annexin family of proteins, can bind to lipid membranes (38, 48). Although the exact physiological function of synexin remains unknown, it has been proposed to act as a Ca²⁺ channel and as a Ca²⁺-activated GTPase, thus regulating intracellular vesicle fusion and membrane trafficking (38, 48-50). Galectin-3 contains no signal sequence for its subcellular localization and is present in the nucleus and cytoplasm. It is also secreted by a mechanism independent of the endoplasmic reticulum-Golgi secretory vesicle pathway. It has been proposed that galectin-3 is targeted to the plasma membrane through an unknown mechanism and secreted through vesicular budding, followed by release from the vesicle. Down-regulation of synexin prevents galectin-3 translocation to the perinuclear membranes and increases galectin-3 secretion. This shows that synexin does not direct galectin-3 exocytosis, but is required for intracellular translocation of galectin-3. Thus, it appears that galectin-3 secretion and intracellular translocation employ two different pathways and that disruption of intracellular galectin-3 localization indirectly promotes the galectin-3 secretion pathway, suggesting a cross-talk between these pathways. Synexin regulates vesicle fusion in a Ca²⁺-dependent manner (38, 48, 49). Interestingly, galectin-3 secretion was shown to be markedly stimulated by the calcium ionophore A23187, which alters homeostasis (51), suggesting that Ca²⁺ may be a messenger for the cross-talk between the galectin-3 transport pathways.

	ACKNOWLEDGEMENTS

We thank Dr. Carl E. Creutz for providing the GST-synexin expression vector, Dr. Kamiar Moin and Linda Mayernik for helping with confocal microscopic analysis, and Mary Ann Krug for helping with preparation of the manuscript.

	FOOTNOTES

^* This work was supported in part by NCI Grant CA64139 from the National Institutes of Health, Department of Defense Grant DAMD17-99-1-9442 from the United States Army, and a Virtual Discovery grant from the Karmanos Cancer Institute (to H.-R. C. K.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Dept. of Pathology, Wayne State University School of Medicine, 540 E. Canfield, Detroit, MI 48201. Tel.: 313-577-2407; Fax: 313-577-9165; E-mail: hrckim@med.wayne.edu.

Published, JBC Papers in Press, February 11, 2002, DOI 10.1074/jbc.M200154200

² R. L. Finley, Jr., unpublished data.

³ K. L. Lavine and R. L. Finley, Jr., unpublished data.

	ABBREVIATIONS

The abbreviations used are: DEVDase, (Asp-Glu-Val-Asp)ase; PBS, phosphate-buffered saline; FITC, fluorescein isothiocyanate; Raf, raffinose; X-gal, 5-bromo-4-chloro-3-indolyl--D-galactopyranoside; GST, glutathione S-transferase.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES