Search tips
Search criteria 


Logo of molcellbPermissionsJournals.ASM.orgJournalMCB ArticleJournal InfoAuthorsReviewers
Mol Cell Biol. 2010 April; 30(7): 1800–1813.
Published online 2010 February 1. doi:  10.1128/MCB.01357-09
PMCID: PMC2838071

Bax Inhibitor 1 Increases Cell Adhesion through Actin Polymerization: Involvement of Calcium and Actin Binding[down-pointing small open triangle]


Bax inhibitor 1 (BI-1), a transmembrane protein with Ca2+ channel-like activity, has antiapoptotic and anticancer activities. Cells overexpressing BI-1 demonstrated increased cell adhesion. Using a proteomics tool, we found that BI-1 interacted with γ-actin via leucines 221 and 225 and could control actin polymerization and cell adhesion. Among BI-1−/− cells and cells transfected with BI-1 small interfering RNA (siRNA), levels of actin polymerization and cell adhesion were lower than those among BI-1+/+ cells and cells transfected with nonspecific siRNA. BI-1 acts as a leaky Ca2+ channel, but mutations of the actin binding sites (L221A, L225A, and L221A/L225A) did not change intra-endoplasmic reticulum Ca2+, although deleting the C-terminal motif (EKDKKKEKK) did. However, store-operated Ca2+ entry (SOCE) is activated in cells expressing BI-1 but not in cells expressing actin binding site mutants, even those with the intact C-terminal motif. Consistently, actin polymerization and cell adhesion were inhibited among all the mutant cells. Compared to BI-1+/+ cells, BI-1−/− cells inhibited SOCE, actin polymerization, and cell adhesion. Endogenous BI-1 knockdown cells showed a similar pattern. The C-terminal peptide of BI-1 (LMMLILAMNRKDKKKEKK) polymerized actin even after the deletion of four or six charged C-terminal residues. This indicates that the actin binding site containing L221 to D231 of BI-1 is responsible for actin interaction and that the C-terminal motif has only a supporting role. The intact C-terminal peptide also bundled actin and increased cell adhesion. The results of experiments with whole recombinant BI-1 reconstituted in membranes also coincide well with the results obtained with peptides. In summary, BI-1 increased actin polymerization and cell adhesion through Ca2+ regulation and actin interaction.

In metastasis, tumor cells migrate from primary tumor sites into the lymphatic or circulatory system and then attach to the basal matrix of the target tissue (16). Cell adhesion and migration contribute to the metastatic process. Adhesion assembly and turnover are highly dynamic, coordinated processes essential for cell migration (16, 26). Adhesions serve as traction points for cell translocation and mediate a network of signaling events that regulate protrusion, contractility, and attachment (16, 29, 30). In migrating cells, protrusions are generated by actin polymerization at the front of the cell (22). Actin exists as monomers (G-actin) and polymers (F-actin), which transform into each other, and the transformation has a major contribution to cell physiology and dynamics. In the cell under physiological conditions, both G- and F-actin contain Mg2+ at the high-affinity binding site. The actin dynamic state contributes to cancer metastasis environments, including that of increased cell adhesion.

The antiapoptotic protein Bax inhibitor 1 (BI-1) was identified through a functional yeast screen designed to select for human cDNAs that inhibit Bax-induced apoptosis (39). BI-1 regulates Ca2+ levels in the endoplasmic reticulum (ER) and cytosol (19) via a C-terminal amino acid sequence of EKDKKKEKK. The antiapoptotic function of BI-1 contributes to the development of cancer and resistance to antitumor therapies (12, 14, 17), but the roles of BI-1 in regulating cell adhesion and actin polymerization are unclear. This study examines the role of BI-1 in cell adhesion through Ca2+ regulation and actin polymerization.


Cell culture.

Human HT1080 fibrosarcoma cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 20 mM HEPES, 100 μg/ml streptomycin, and 100 units/ml penicillin. HT1080 cells were stably transfected with the pcDNA3 (neomycin resistance [Neo]) vector; the plasmid pcDNA3-BI-1-HA, expressing wild-type hemagglutinin [HA]-tagged BI-1; a plasmid expressing HA-tagged BI-1 with the C terminus deleted (CΔBI-1); or a plasmid expressing the HA-tagged L221A, L225A, or L221/225A BI-1 actin binding site mutant by using Superfect transfection reagent (Qiagen, Hilden, Germany). The cells were then cultured for 3 weeks in 1 mg/ml G418 (Invitrogen, CA).


Alexa Fluor 488-conjugated phalloidin and pyrene maleimide were purchased from Molecular Probes (Carlsbad, CA). The antibody against HA antigen was purchased from Cell Signaling Technologies (Beverly, MA). DMEM, FBS, trypsin, and other tissue culture reagents were supplied by Life Technologies, Inc. (Grand Island, NY). Bicinchoninic acid (BCA) protein assay reagents were obtained from Pierce Biotechnology (Rockford, CA). All other chemicals were at least of analytical grade and were purchased from Sigma Chemical Company (St. Louis, MO).

Electron microscopy.

Neo and BI-1 cells were collected and prepared for transmission electron microscopy (TEM) by fixation in a phosphate-buffered solution (Sorensen's phosphate, pH 5.8) of 2.5% glutaraldehyde with 0.15% sucrose and 2% mannitol to maintain proper osmosis for 12 to 24 h at 4°C. Cells were postfixed in 2% osmium tetroxide in the same buffer for 1 h. After being rinsed, the cells were exposed to 2% uranyl acetate (aqueous) and subsequently embedded in 2% agar. Samples cut from the agar blocks were dehydrated in a graded ethanol series, infiltrated with Spurr's resin, embedded, and cured at 70°C for 24 h. Ultrathin sections were cut and stained with uranyl acetate and Reynold's lead citrate prior to observation with a Zeiss EM10 transmission electron microscope.


Cells grown on coverslips were fixed with 4% paraformaldehyde and permeablized with 0.5% Triton X-100 in phosphate-buffered saline (PBS) for 15 min at room temperature. After being blocked with 3% bovine serum albumin in PBS for 15 min, the cells were incubated with Alexa Fluor 488-conjugated phalloidin (1:40 dilution) for F-actin staining and viewed using an Axioplan fluorescence microscope (Carl Zeiss MicroImaging, Thornwood, NY).

Cell adhesion assays.

Neo and BI-1-overexpressing HT1080 cells (104 cells/cm2) were trypsinized and washed twice with 1% FBS medium. Cells were held in suspension in 1% FBS medium for 30 min and then seeded into 12-well plates (n = 7) in 1% FBS medium. Cell samples were gently washed twice with PBS to remove nonadherent cells. The adherent cells were fixed with 3.7% formaldehyde, stained with 0.05% crystal violet, and washed three times with deionized water. The dye was eluted with 80% methanol, and the absorbance at 570 nm was measured.


Cells were seeded onto six-well plates for 24 h and then subjected to serum starvation for 24 h. After drug treatment for the times indicated in the figures, treated cells were harvested by being washed twice with ice-cold PBS on ice. For the preparation of whole-cell lysates, cells were lysed on ice by the addition of RIPA lysis buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 0.25% sodium deoxycholate, 1% NP-40, 1 mM EDTA, 0.1% sodium dodecyl sulfate [SDS]) plus protease inhibitor cocktail set III and phosphatase inhibitor cocktail set II (EMD Biosciences, La Jolla, CA) directly onto the cells. Cell lysates were then transferred into microtubes, incubated for 30 min on ice, and centrifuged at 12,000 rpm for 10 min at 4°C, and supernatants were collected to obtain protein extracts. Protein extracts were then added to sample buffer, boiled in a water bath for 5 min, and stored at −20°C until use. Samples of 10 μg of extracted proteins were run on polyacrylamide gel and transferred onto a nitrocellulose membrane, which was then blocked with Tris-buffered saline solution containing 0.05% Tween 20 for 30 min at room temperature. The blots were then probed overnight at 4°C with the relevant antibodies, washed, and probed again with species-specific secondary antibodies coupled to horseradish peroxidase (GE Healthcare, Piscataway, NJ). Chemiluminescence reagents (GE Healthcare) were then added for blot analysis.


Cells were lysed with IP buffer (10 mM Tris [pH 7.4], 150 mM NaCl, 1% Triton X-100, 1% NP-40) supplemented with protease and phosphatase inhibitors. Cell lysates were then transferred into microtubes, incubated for 30 min on ice, and subjected to centrifugation at 12,000 rpm for 10 min at 4°C. Supernatants were collected, and protein concentrations were measured by the Bradford assay. Protein was immunoprecipitated with the antibodies indicated below at 4°C overnight and then incubated with protein G-agarose (Roche Applied Science, Indianapolis, IN) at 4°C for 3 h. After being washed five times with IP buffer, the immunoprecipitates were analyzed by immunoblotting for protein identification.

Silver staining.

After electrophoresis, the gels were silver stained according to the instructions of the stain manufacturer (Amersham Pharmacia). Briefly, the gels were fixed (in a mixture of 10% glacial acetic acid and 40% methanol), sensitized (in a mixture of 0.5% glutardialdehyde, 0.2% sodium thiosulfate, 5.6% sodium acetate, and 30% ethanol), rinsed with deionized water, and then silver stained (with a solution of 0.25% silver nitrate and 0.01% formaldehyde). The protein spots in the gels were developed with a solution of 0.25% sodium carbonate and 0.01% formaldehyde.


Silver-stained proteins were destained with chemical reducers to remove the silver as described previously, with some modifications (14). Potassium ferricyanide and sodium thiosulfate were prepared as two stock solutions of 30 mmol/liter potassium ferricyanide and 100 mmol/liter sodium thiosulfate dissolved in water. A working solution was prepared by mixing the stock solutions at a 1:1 ratio prior to use. After interesting protein spots were excised from the gel, 30 to 50 liters of the working solution was added to cover the gel, which was occasionally subjected to a vortex. The stain intensity was monitored until the brownish color disappeared, and the gel was then rinsed a few times with water to stop the reaction. Next, 200 mmol/liter ammonium bicarbonate was added to cover the gel for 20 min and then discarded. The gel was subsequently cut into small pieces, washed with water, and dehydrated with changes of acetonitrile until the gel pieces turned opaque white. The gel pieces were dried in a vacuum centrifuge for 30 min.

Image analysis.

The silver-stained gel was scanned with LabScan software using an ImageScanner instrument (Amersham Biosciences), and the data were digitized and analyzed using ImageMaster 2D (Amersham Biosciences).

Trypsin digestion of in-gel proteins.

Enzymatic digestion was performed as described previously (38). In brief, digestion was performed with a mixture of 5 to ~10 ng/liter trypsin and 50 mmol/liter ammonium bicarbonate incubated overnight at 37°C. Following enzymatic digestion, the resultant peptides were extracted three times with 10 to ~20 liters of 5% trifluoroacetic acid in 50% acetonitrile and dried using a vacuum centrifuge for 30 min.

Identification of proteins.

Dried samples were sent to the Korea Basic Science Institute, Daejeon, South Korea, for matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry analysis with a Voyager-DE PRO system and for electrospray ionization quadrupole time of flight (ESI-Q-TOF) mass spectrometry analysis. Peptide mass fingerprinting was analyzed by MALDI-TOF mass spectrometry. The results of peptide mass fingerprinting were identified using the PeptIdent program ( Additionally, peptide sequencing was performed by using an ESI-Q-TOF mass spectrometer to confirm the identification by MALDI-TOF mass spectrometry.

Analysis of Triton-insoluble cytoskeletal fraction.

Cells were lysed in Triton X-100 lysis buffer (0.5% Triton X-100, 0.3 M sucrose, 100 mM NaCl, 3 mM MgCl2) supplemented with protease and phosphatase inhibitors. Cell lysates were then transferred into microtubes, incubated for 10 min on ice, and centrifuged at 12,000 rpm for 10 min at 4°C, and supernatants were removed. The pellets were washed with Triton X-100 lysis buffer and then resuspended in SDS buffer (1% SDS, 2 mM EDTA, 2 mM EGTA, 20 mM Tris-HCl [pH 7.4]) supplemented with protease and phosphatase inhibitors. The Triton-insoluble fraction was boiled for 10 min and then chilled on ice. Protein levels were determined by using the BCA protein assay kit (Pierce, Rockford, IL). Samples of 10 μg of protein were analyzed by immunoblotting.

Cell culture and transfection for electrophysiological recordings.

HEK 293 cells were maintained in DMEM with 10% FBS at 37°C in a humidified atmosphere under 5% CO2. For transient transfection, cells were grown to ~70% confluence in six-well plates. The next day, cells were transfected with 0.4 μg/well of a pcDNA construct expressing green fluorescent protein (GFP)-BI-1 by using Effectene transfection reagent according to the protocol of the manufacturer (Qiagen, Valencia, CA). Cells were cotransfected with CD8 (Dynabeads M-450 CD8; Dynal, Oslo, Norway) as an expression marker. After 18 to 24 h, cells were trypsinized and replated onto poly-l-lysine-coated coverslips. The BI-1-expressing cells were identified by incubating the cells with CD8 beads or by detecting the expression of GFP.

Electrophysiological recordings.

Whole-cell patch-clamp recordings were performed with an EPC-10 patch-clamp amplifier (HEKA Elecktronik, Lambrecht/Pfalz, Germany). Data were acquired using the Pulse program (HEKA Elecktronik). Extracellular solutions contained 140 mM NaCl, 10 mM HEPES, 2 mM CaCl2, 1 mM MgCl2, 10 mM glucose, and 5 mM KCl and were adjusted to pH 7.4 with NaOH. The Ca2+-free extracellular solution contained 3 mM MgCl2 instead of CaCl2. Pipette solutions consisted of 135 mM CsCl, 10 mM HEPES, 10 mM EGTA, 5 mM Mg-ATP, 5 mM MgCl2, and 10 mM glucose and were adjusted to pH 7.3 with CsOH. When the pipettes were filled with the pipette solution, the resistance of the pipettes was 4 to 5 MΩ and the series resistance was compensated for (>80%). The holding potential was −60 mV. The recording chamber (volume, approximately ~500 μl) was perfused via gravity at a rate of 2 to 3 ml/min at room temperature. In this study, 10 mM CaCl2 was applied via a glass puffer pipette (1-μm tip diameter; pressure, ≈10 lb/in2 [69 kPa]) using a pneumatic PicoPump (PV830; World Precision Instruments, Stevenage, Herts, United Kingdom). The tip of the puffer pipette was positioned downstream from the cell with respect to the direction of the flow of the Ca2+-free extracellular recording solution and temporarily repositioned to a point about 15 μm from the cell only for the period of application.

Intracellular Ca2+ measurements.

The fluorescent calcium indicator Fura-2AM {1-[2-(5-carboxyoxazol-2-yl)-6-aminobenzofuran-5-oxy]-2-(2-amino-5-methylphenoxy)-ethane-N,N,N′,N′-tetraacetic acid pentaacetoxymethyl ester; Molecular Probes, Eugene, OR} was used to measure changes in intracellular (cytosolic) free Ca2+. Cells were plated onto glass-bottomed perfusion chambers that were mounted onto the stage of an inverted microscope (Nikon Eclipse TE2000) and incubated with Fura-2AM (6 μM) for 30 min at room temperature in Hanks’ balanced salt solution (HBSS). After being loaded, cells were washed three times in isotonic buffer without Ca2+ (KH buffer: 132 mM NaCl, 5 mM KCl, 10 mM dextrose, 10 mM HEPES, 1.05 mM MgCl2). Cells were then promptly treated with various agents, including thapsigargin, ionomycin, ryanodine, histamine, and ATP. Changes in the intracellular Ca2+ concentration ([Ca2+]i) were determined as ratios corresponding to 340/380-nm excitation wavelengths (emission wavelength, 512 nm) by using an integrated spectrofluorometer (Photon Technology International, Birmingham, NJ). Ca2+ concentrations were calculated using the following equation: [Ca2+]i = Kd(F380max/F380min)(RRmin)/(RmaxR), where Kd is the dissociation constant of Fura-2 to calcium, F380max/F380min is the ratio of fluorescence emission intensity at 380-nm excitation in Ca2+-depleting (F380max) and Ca2+-saturating (F380min) conditions, and Rmin and Rmax are the minimum and maximum F340/F380 ratios, respectively. A Kd of 229 nM for the binding of calcium to Fura-2AM was assumed. Rmax and Rmin were determined in each experimental group by the consecutive addition of 30 μM digitonin (for Rmax) and 50 mM EGTA (for Rmin).

Actin preparation.

Ca-ATP-G-actin from back and leg muscles of rabbits was prepared by the method of Spudich and Watt (32) and stored in buffer containing 5.0 mM Tris-HCl, 0.2 mM CaCl2, 0.2 mM ATP, and 0.5 mM β-mercaptoethanol, pH 8.0 (Ca-ATP-G-buffer). Mg-ATP-G-actin, which is the physiological form of G-actin, was obtained by incubating Ca-ATP-G-actin with 0.2 mM EGTA and 0.1 mM MgCl2 at room temperature for 5 min. Mg-ATP-G-actin (≤10 μM) was used within 2 h after preparation. Mg-ATP-G-actin was diluted in buffer containing 5 mM Tris-HCl, 0.1 mM MgCl2, 0.2 mM EGTA, 0.2 mM ATP, and 0.5 mM dithiothreitol (DTT), pH 8.0 (Mg-ATP-G-buffer), for further treatments. The concentration of unlabeled skeletal muscle α-G-actin was determined spectrophotometrically using the extinction coefficient of 1% solution at 290 nm (E1%290) of 11.5 cm−1. The optical density of actin was measured in the presence of 0.5 M NaOH, which shifts the maximum absorbance from 280 to 290 nm. The molecular mass of skeletal actin was assumed to be 42 kDa.

Chemical modification.

Labeling of actin at Cys-374 with pyrene maleimide was carried out according to the method of Kouyama and Mihashi (21) with some modifications. Ca-ATP-G-actin was filtered through a PD-10 column equilibrated with a buffer containing 5.0 mM Tris-HCl, 0.2 mM CaCl2, and 0.2 mM ATP, pH 8.0. After filtration, actin (1.0 mg/ml) was polymerized by 2.0 mM MgCl2 and 100 mM KCl at room temperature for 30 min and reacted with pyrene maleimide (16 μg/ml) on ice for 1 h. The reaction was terminated with 1.0 mM DTT. The labeled F-actin was centrifuged at 38,000 rpm for 2 h, and then the pellet was suspended in Ca-ATP-G-buffer and depolymerized for over 36 h at 4°C. Finally, actin was centrifuged again at 38,000 rpm for 2 h. The supernatant contained the purified pyrene-labeled Ca-ATP-G-actin. The concentration of modified actin was determined by the Bradford procedure (4) using unmodified actin as a standard. The extent of labeling, which was measured by using the pyrene extinction coefficient (E344) of 22 mM−1, was ~100%.

Fluorescence and light-scattering measurements.

The time course of pyrene-labeled actin polymerization was monitored by measuring the fluorescence increase (with 365-nm excitation and 386-nm emission wavelengths) with a spectrofluorometer (Photon Technology Industries, South Brunswick, NJ). Light scattering was also measured with a spectrofluorometer, with both excitation and emission wavelengths adjusted to 450 nm.

Expression, purification, and reconstitution of recombinant BI-1 protein.

Recombinant BI-1 was expressed, purified, and reconstituted in liposomes as described previously (19). The mutant BI-1 proteins including C-terminal deletions and mutations L221A, L225A, and L221A/L225A were constructed using a QuikChange XL site-directed mutagenesis kit (Stratagene, La Jolla, CA). The mutants were expressed and purified by the same method used for wild-type BI-1. To compare Ca2+ channel activities between intact BI-1 and the mutants, pH decrease-induced Ca2+ release from the reconstituted proteoliposomes was also measured as described previously (19).

Statistical analysis.

Results are presented as the means ± standard errors of the means of n experiments, and paired and unpaired Student t tests were applied where appropriate. Origin software (Microcal, Northampton, MA) was used for statistical calculations.


BI-1 overexpression increases cell adhesion.

BI-1-overexpressing cells show highly metastatic phenotypes. Electron microscopy analysis showed that BI-1 cells formed lamellipodia (Fig. (Fig.1A).1A). At 24 h after plating, BI-1 cells had highly developed edges and ruffles at the periphery, in contrast with Neo cells. BI-1 cells also showed faster adhesion to slides coated with poly-l- or poly-d-lysine, with adhesion in less than 40 min versus over 1 h for Neo cells (Fig. (Fig.1B).1B). CΔBI-1 mutant cells showed a pattern similar to that of Neo cells. All of the attached cells excluded trypan blue and propidium iodide and showed light refraction under interference microscopy, which is typical of live cells (Fig. (Fig.1C).1C). No evidence of apoptosis, such as chromatin condensation and cell shrinkage, could be seen (data not shown). Since focal adhesion kinase (FAK) plays central roles in adhesive interactions by functioning as a scaffold for focal adhesion components such as paxillin (31), the expression patterns of FAK and paxillin in Neo, BI-1, and CΔBI-1 cells were observed. The relative levels of expression of the adhesion-associated proteins in BI-1 cells were higher than those in Neo or CΔBI-1 cells (Fig. (Fig.1D1D).

FIG. 1.
BI-1 overexpression increases cell adhesion. (A and B) For electron microscopy analyses, Neo, BI-1, and CΔBI-1 cells were trypsinized and plated onto dishes coated with lysine (A), and adherent cells were counted (B). *, P < 0.05 ...

Actin interacts with BI-1.

We paired immunoprecipitation using anti-HA to pull down the expression plasmid with protein identification analysis using liquid chromatography-mass spectrometry and identified actin (Fig. (Fig.2A)2A) and γ-actin (Fig. (Fig.2B)2B) interacting with BI-1. In a bioinformatics approach, the solutions for structures of actin complexes with actin binding proteins (ABPs), including actin-gelsolin, actin-ciboulot, actin-kabiramide C, and actin-vitamin D binding protein (10), indicated low levels of sequence similarity for interaction sites but consistent interaction with the hydrophobic pocket of actin and the exposed hydrophobic side chains of the alpha-helices of ABPs containing LXXXL or LXXXI motifs (data not shown). The secondary structure of BI-1 was predicted by PSIPRED (25). BI-1 contains a conserved L221XXXL225 motif in the terminal alpha-helix (data not shown). Introducing the mutation L221A, L225A, or both disrupted BI-1 interaction with γ-actin, but the deletion of the EKDKKKEKK C-terminal sequence (yielding CΔBI-1) did not (Fig. (Fig.2C),2C), suggesting that the C-terminal motif is not an actin binding domain.

FIG. 2.
γ-Actin interacts with BI-1. (A and B) Immunoprecipitation (IP) of BI-1 and Neo cell extracts was performed with IgG or HA antibodies. SDS-PAGE and subsequent silver staining (A) and immunoblotting with anti-γ-actin or anti-HA antibody ...

BI-1 induces actin polymerization and cell adhesion via the C terminus and actin binding sites.

Actin polymerization is important for cell adhesion. Mutation of the actin binding site in BI-1 reduced the amount of polymerized actin in the membrane fraction to a level similar to that in the membrane fraction from CΔBI-1 cells (Fig. (Fig.3A).3A). For all cell groups, actin labeling with fluorescent phalloidin showed parallel actin cables in the cell body and actin-rich rims at the edges of lamellipodial protrusions. Actin cables (stress fibers) in the BI-1 cells formed densely packed parallel arrays that traversed the entire cell (Fig. (Fig.3B).3B). Mutant BI-1 cells showed lower levels of phalloidin fluorescence. Furthermore, differences in actin polymerization can influence adhesion ability. Cells expressing CΔBI-1 and the L221A, L225A, and L221A/L225A mutants showed lower adhesion activity than BI-1-expressing cells (Fig. (Fig.3C;3C; also data not shown). Consistently, the levels of expression of FAK and paxillin were also lower in all the BI-1 mutant cell groups than in BI-1 cells (Fig. (Fig.3D).3D). Consistent results were seen with BI-1-expressing mouse embryo fibroblasts (MEF), with cells expressing endogenous BI-1 (BI-1+/+ cells) showing larger amounts of polymerized actin (Fig. (Fig.4A)4A) and more extensive actin polymerization (Fig. (Fig.4B)4B) than BI-1−/− cells. Consistently, BI-1+/+ MEF cells had increased adhesion ability compared with BI-1−/− MEF cells (Fig. (Fig.4C;4C; also data not shown). The levels of expression of FAK and paxillin in BI-1+/+ MEF cells were also higher than those in BI-1−/− MEF cells (Fig. (Fig.4D4D).

FIG. 3.
BI-1 induces actin polymerization via the C terminus and actin binding sites. (A) Neo, BI-1, CΔBI-1, and L221A, L225A, and L221A/L225A mutant cells were fractionated into Triton-soluble and -insoluble components, and immunoblotting was performed ...
FIG. 4.
Endogenous BI-1 induces actin polymerization and cell adhesion. (A) BI-1+/+ and BI-1−/− MEF cells were fractionated and subjected to immunoblotting as described in the legend to Fig. Fig.3A,3A, and then the F-action/G-actin ...

We also examined the role of endogenous BI-1 in HT1080 cells by transfecting HT1080 cells with BI-1 small interfering RNA (siRNA). siRNA inhibited BI-1 expression (Fig. (Fig.4E).4E). Compared with nonspecific siRNA, BI-1 siRNA decreased actin movement into the Triton-insoluble membrane fraction, yielding a relatively low ratio of F-actin to G-actin (Fig. (Fig.4F).4F). BI-1 siRNA also decreased cell adhesion (Fig. (Fig.4G)4G) and the levels of FAK and paxillin expression (Fig. (Fig.4H4H).

BI-1-releasable Ca2+ and the presence of actin binding sites are required for actin polymerization.

BI-1 has a Ca2+ channel function (19), and continuous Ca2+ leakage can generally enhance actin polymerization (8, 27). When cells were treated with thapsigargin in order to measure [Ca2+]i, BI-1 cells showed lower [Ca2+]i than Neo cells, also suggesting a lower intra-ER concentration of Ca2+ ([Ca 2+]ER) (Fig. (Fig.5A).5A). Expectedly, the CΔBI-1 cells showed high [Ca2+]i, similar to that in Neo cells (Fig. (Fig.5A);5A); however, cells expressing L221A, L225A, and L221A/L225A mutants showed responses similar to that of the BI-1 cells (Fig. (Fig.5B),5B), meaning that the C terminus is essential for Ca2+ channel activity but the actin binding site is not. Although supplying external Ca2+ after thapsigargin treatment evoked [Ca2+]i increase in all groups, the amplitudes of [Ca2+]i changes differed between Neo and BI-1 cells, which may be due to the activation of a store-operated Ca2+ entry (SOCE) pathway by the reduced [Ca2+]ER in BI-1 cells. It was therefore of interest to evaluate the effect of BI-1 and its mutants on consequent SOCE. To evaluate SOCE, changes in intracellular Ca2+ concentrations were monitored in cells in which Ca2+ stores were depleted with thapsigargin, a sarco/ER Ca2+-ATPase (SERCA) inhibitor, in the presence of Ca2+-free extracellular solution, after which Ca2+ was added. Expectedly, the BI-1 cells exhibited highly increased Ca2+ entry through SOCE channels, which may also conceivably result from an increased driving force for Ca2+ influx from plasma membrane transporters (Fig. (Fig.5C).5C). Ca2+ entry was inhibited in all cells expressing BI-1 mutant forms, including actin binding site mutants (Fig. (Fig.5D),5D), demonstrating differences in the Ca2+ release pattern among the BI-1 mutants (Fig. (Fig.5B).5B). 2-Aminoethoxydiphenyl borate (2-APB; 20 μM), a SOCE inhibitor, abrogated Ca2+ entry, confirming the activation of SOCE in BI-1 cells (Fig. (Fig.5E).5E). In the presence of 2-APB, however, a small Ca2+ peak was observed in BI-1 cells but not in Neo cells. There is the possibility of another source of Ca2+ influx through a Ca2+ leakage/channel-like function of BI-1. Thus, we tested whether BI-1 is expressed in the plasma membrane. HEK 293 cells were transiently transfected with GFP-BI-1-pcDNA3. Although BI-1 is expressed mainly in the ER membrane (39), the expression of BI-1 in the plasma membrane was also confirmed (Fig. (Fig.6A).6A). Pure fractionation was confirmed by immunoblotting with antibodies specific for the plasma membrane (Fas ligand [Fas-L]), the cytoplasm (procaspase-3), mitochondria (Hsp60), and the ER (calreticulin). To determine whether BI-1 has the property of being an ion channel, we recorded the ionic currents from HEK 293 cells transfected with BI-1. Repetitive puffing of 10 mM Ca2+ under Ca2+-free conditions induced inward currents into HEK 293 cells transfected with GFP-BI-1-pcDNA3 (Fig. (Fig.6B).6B). However, we did not observe any change in HEK 293 cells transfected with GFP-pcDNA3 after puffing of 10 mM Ca2+. This result showed that BI-1 is also expressed in the plasma membrane and that it could act as a Ca2+ leak channel in BI-1-overexpressing HEK 293 cells. Together, these data suggested that both the actin binding affinity and the Ca2+ entry function are required for actin polymerization and cell adhesion. To investigate the effect of endogenous BI-1 on the Ca2+ dynamic status, Ca2+ entry into BI-1+/+ and BI-1−/− MEF cells was examined. After the treatment of cells with thapsigargin in the presence of Ca2+-free extracellular solution and the addition of Ca2+, Ca2+ influx into BI-1−/− MEF cells was significantly reduced (Fig. (Fig.5F),5F), showing consistency among the BI-1-overexpressing cells (Fig. (Fig.5C5C).

FIG. 5.
The release of Ca2+ via BI-1 regulates actin polymerization. (A) Neo and BI-1 cells were loaded with Fura-2AM, cultured in Ca2+-free medium, and incubated with 1 μM thapsigargin. Individual cells (n = 16) were imaged, and ...
FIG. 6.
BI-1 has a Ca2+ leakage channel function on the plasma membrane. (A) HEK 293 cells were transiently transfected with GFP-pcDNA3 or GFP-BI-1, and then the expression of BI-1 was evaluated with anti-GFP or antiactin antibody. Plasma membranes were ...

Actin polymerization increases cell adhesion.

Actin polymerization can influence cell adhesion (3, 29). The physiological consequences of BI-1-induced actin polymerization in relation to cell adhesion need to be studied. Jasplakinolide induces polymerization of monomeric actin into amorphous masses (36). In order to confirm the effect of jasplakinolide on actin polymerization, the plasma membrane fraction from HT1080 cells cultured with or without the polymerization agent was treated with 1% (wt/vol) Triton X-100, which solubilizes membrane lipids and integral proteins, to obtain a Triton X-100-insoluble plasma membrane pellet enriched with F-actin (Fig. (Fig.7A)7A) (35). In the presence of the polymerization agent, the concentration of actin in the Triton-insoluble membrane fraction was high and cell morphology revealed by phalloidin staining also showed an increase in actin polymerization (Fig. (Fig.7B).7B). Cell adhesion also increased in the presence of jasplakinolide (Fig. (Fig.7C),7C), consistent with the results of previous studies (36).

FIG. 7.
Actin polymerization increases cell adhesion. (A) HT1080 cells were treated with or without jasplakinolide (Jas). Triton-soluble and -insoluble fractions were prepared, and SDS-PAGE and immunoblotting were performed. *, P < 0.05 for comparison ...

An actin binding site is required for actin polymerization in vitro.

We needed to check the consistency of BI-1 function relative to actin polymerization in vivo and in vitro. First, in order to confirm the role of BI-1 in actin polymerization in vitro, actin polymerization in the presence of various concentrations of BI-1 protein was measured. An increase in the amount of reconstituted BI-1 increased polymerization (Fig. (Fig.8A).8A). For in vivo BI-1-induced actin polymerization, one of the important factors for cell adhesion, dynamic Ca2+ movement and an actin binding site are required (Fig. (Fig.33 and and5).5). To study the two requirements in vitro, experiments with whole BI-1 proteins and mutant peptides with 2-, 4-, and 6-residue C-terminal truncations and the mutations L221A, L225A, and L221A/L225A were performed. Figure Figure8B8B shows that until four C-terminal residues were deleted, the polymerization induced by the protein was similar to that induced by native BI-1 protein. However, polymerization was significantly decreased by the 6-residue deletion. The whole BI-1 proteins with mutations in the actin binding site also produced results very similar to those obtained with L221A, L225A, and L221A/L225A mutant peptides, as all the mutations decreased actin polymerization compared to that by native BI-1 protein and the protein with the double mutation lacked activity (Fig. (Fig.8C).8C). As a control experiment, the mutant and native proteins were reconstituted in liposomes containing encapsulated Ca2+ and their proton-induced Ca2+ release activities were measured as described previously (19). The Ca2+ efflux activity was marginally decreased with the mutation of the actin binding site compared to that of native BI-1 (Fig. (Fig.8D).8D). This result suggests that the actin binding site mutations do not influence the Ca2+ channel activity of the whole protein and possibly the topology of the whole protein in the membrane, although the precise conformation of BI-1 protein in the membrane was not revealed. To clarify the BI-1 protein results, the 9-amino-acid peptide EKDKKKEKK and the 17-amino-acid peptide LMMILAMNEKDKKKEKK were added to purified Mg-ATP-G-actin, together with 100 mM NaCl to maintain physiological ionic strength. The wild-type LMMILAMNEKDKKKEKK peptide (WT-17), containing actin binding sites (L221 and L225), strongly activated actin polymerization, even after the deletion of its four charged C-terminal residues (yielding the Δ4 mutant). However, the deletion of 6 residues from the C terminus of WT-17 (generating the Δ6 mutant) remarkably reduced activation. The 9-residue peptide EKDKKKEKK, containing the C-terminal motif, did not activate polymerization (Fig. (Fig.9A).9A). L221A, L225A, and L221A/L225A mutant 17-mer peptides were tested for Mg-ATP-G-actin polymerization in comparison with WT-17. The actin-polymerizing abilities of these peptides decreased in the following order: WT-17, the L221A peptide, the L225A peptide, and the L221A/L225A peptide (Fig. (Fig.9B).9B). We also studied the effects of scrambling the residues in WT-17 on actin polymerization and found that the scrambled peptides (both LNEKLKDMKIKMKAEKM, in which L221 and L225 remained in the original positions, and KEMKDILKENMKLKMAK, which was completely scrambled) essentially did not activate polymerization (Fig. (Fig.9C).9C). Taken together, the results suggest that the actin binding site-containing 11-residue sequence LMMILAMNEKD is essential for actin interaction and that the addition of two lysines to the C terminus of the sequence further increases the activity, indicating that these residues may have a supporting role in the process and that the highly positively charged C-terminal motif without the actin binding site is not directly involved in the polymerization of actin. The Mg-F-actin-bundling activities of these peptides were also examined. WT-17 bundled actin filaments, and the mutated L221A peptide also had slight bundle formation activity; however, the L225A and L221A/L225A mutants did not induce bundle formation (Fig. (Fig.9D).9D). The results suggest that L221 is the most important residue in WT-17 for polymerizing and bundling actin, at least in the in vitro system.

FIG. 8.
Effects of C-terminal truncations and mutations in the acting binding sites of BI-1 protein. (A to C) Actin polymerization, measured as an increase in pyrene fluorescence, was induced in Mg-ATP-G-buffer, pH 7.2, by using 6.5 μM BI-1 proteins reconstituted ...
FIG. 9.
The actin binding site of the C-terminal WT-17 peptide of BI-1 is required for actin polymerization and bundling in vitro. Polymerization, monitored as an increase in pyrene fluorescence, was induced in Mg-ATP-G-buffer, pH 7.2. The arrows indicate the ...

A 17-amino-acid BI-1 peptide increases actin polymerization and cell adhesion.

To evaluate the 17-amino-acid peptide containing actin binding sites and a C-terminal motif, LMMILAMNEKDKKKEKK, the peptide was introduced into HT1080 human fibrosarcoma cells. The expression of γ-actin in both the Triton X-100-insoluble pellet (predominantly F-actin) and the Triton X-100-soluble supernatant (predominantly G-actin) was detected by Western blotting. Compared with the control peptide, the WT-17 peptide increased actin movement into the Triton-insoluble membrane fraction (Fig. 10A). By densitometric analysis, a higher ratio of F-actin to G-actin in the presence of the peptide than in the absence of the peptide was detected (Fig. 10A, bottom). Consistently, the cell morphology revealed by phalloidin staining was clearly different in the presence of the WT-17 peptide. The fluorescence was highly increased with the increase in polymerization (Fig. 10B). In addition, cell adhesion was also remarkably increased when cells were incubated with WT-17 (Fig. 10C), showing consistency in the roles of the actin binding sites and C-terminal motif in the peptide and the recombinant BI-1 protein (Fig. (Fig.88 and and99).

FIG. 10.
The WT-17 BI-1 peptide increases actin polymerization and cell adhesion. (A) The BI-1 peptide LMMILAMNEKDKKKEKK (17Peptide) was added to HT1080 cells, Triton-soluble (G-actin) and -insoluble (F-actin) fractions were prepared, immunoblotting with anti-γ-actin ...


Our findings suggest a novel role for BI-1 in regulating actin polymerization and cell adhesion. Throughout this study, cell adhesion was transiently faster among cells expressing wild-type BI-1 than among cells expressing BI-1 mutants, but all groups showed the same adhesion percentage after a plating time of 90 min (Fig. (Fig.1B1B and and3C).3C). BI-1 knockout and knockdown cell adhesion data were also similar (Fig. 4C and G). In the early stage of adhesion after plating, the presence of BI-1 may be an initiator for adhesion, probably through actin polymerization. Consistently, the expression of BI-1 correlates with a specific stage of cancer progression in small-cell lung carcinoma patients: lymph node metastasis (23a). However, the prognosis is independent of the expression of BI-1 (23a). The role of BI-1 may be to initiate the specific stage of cancer progression through the enhancement of adhesion, especially early adhesion efficiency.

Cell adhesion can be greatly affected by signaling molecules other than BI-1 (15, 24). Rac and other integrin complex molecules are examples. Cell adhesion efficiency is also dependent on the amount of activated Rac, as microinjection with constitutively active Rac increases cell adhesion (24). Rac activation also induces the formation of peripheral focal complexes containing focal adhesion proteins (13, 28). Integrin clustering thus seems to be required for spatial regulation of lamellipodium extension and perhaps for stabilization of the lamellipodium in cell motility (7).

The actin cytoskeleton is a primary determinant of tumor cell motility and metastatic potential. Motility and metastasis are thought to be regulated, in large part, by actin polymerization and its cofactor: Ca2+ mobilization (20). BI-1 reduces [Ca2+]ER, as did the actin binding site mutants, but CΔBI-1 cells had normal [Ca2+]ER. Permeability for Ca2+ was not different in cells expressing BI-1 and actin binding site mutants (data not shown), but actin polymerization and cell adhesion were both inhibited in CΔBI-1 and L221A/L225A BI-1 mutant cells (Fig. 3A and B). The continuous leakage of Ca2+ from the ER and the high degree of actin polymerization in BI-1 cells indicate that Ca2+, most probably indirectly through ABPs, can influence actin polymerization and bundling. BI-1 is localized mainly on the ER membrane but interacts with actin to affect actin polymerization and cell adhesion. The low [Ca2+]ER in BI-1 cells may regulate voltage-dependent Ca2+ influx by activating a store-operated depolarizing current (37). Changes in the distribution and/or activities of inositol triphosphate (IP3) receptors and ryanodine receptors may underlie these changes in Ca2+ release. Maintenance of mobilizable Ca2+ in the ER requires high nutrient concentrations (33, 34) to counteract Ca2+ leakage (6). Whereas high-affinity uptake supplies Ca2+ to the releasable IP3-bound pool at basal [Ca2+]i concentrations (2, 37), low-affinity uptake with leakage allows the ER to function as a “passive” Ca2+ buffer at elevated [Ca2+]i (6, 11).

In BI-1 cells, a highly dynamic Ca2+ status probably enables the ER to exert this dual function and regulates downstream actin polymerization and cell adhesion. Actin polymerization also can enhance Ca2+ release, whereas actin depolymerization decreases Ca2+ release through plasma membrane voltage-dependent Ca2+ channels (18). In this study, the overexpression of BI-1 itself was sufficient to activate SOCE. This increase in Ca2+ entry into BI-1 cells is likely to be initiated by modest Ca2+ store depletion, together with SOCE activation due to actin polymerization, because CΔBI-1 cells with low levels of permeability for Ca2+ (19) and the cells expressing actin binding site mutants inhibited BI-1-induced Ca2+ entry (Fig. (Fig.5E).5E). In addition, another source of Ca2+ entry into BI-1 cells, but not Neo cells, in the presence of a SOCE inhibitor was observed (Fig. (Fig.5E),5E), although this finding does not rule out the possibility of a Ca2+ leakage channel role for plasma membrane-localized BI-1. The active site of BI-1 for diverse BI-1-induced functions, including cell protection and the maintenance of permeability for Ca2+, has already been recognized as the C-terminal motif EKDKKKEKK.

In this study, another site of BI-1 is suggested as part of the functional sequence—the actin binding motif (the segment from L221 to D231, containing L221 and L225). BI-1-induced Ca2+ release and its effect on actin dynamics, together with the interaction of BI-1 with actin, explain the BI-1-induced increase in cell adhesion, which is probably linked to cancer metastasis mechanisms in the BI-1 system. Many different signaling pathways, including growth factors, neurotransmitters, and cell adhesion molecules, affect both cytoskeletal dynamics and cellular calcium homeostasis (1, 23). Our findings suggest the possibility of reciprocal functional relationships between Ca2+ and the cytoskeleton in regulating cell adhesion. Similarly, Ca2+ influx into cells induced rapid polymerization of actin and filopodial extension (23). Reciprocally, actin polymerization may enhance Ca2+ release from the Ca2+ storage sites, providing a mechanism for local amplification of the Ca2+ signal and hence increased cell adhesion.

The 17-mer C-terminal peptide of BI-1 was shown to polymerize actin very efficiently. Since this peptide also contains the C-terminal motif with six positively charged lysine residues, one may speculate that its interaction with the negatively charged actin is due to the highly positively charged nature of the C-terminal segment of the peptide. However, we can exclude this possibility because even after the deletion of four charged residues from the C terminus, the peptide retains its full polymerizing activity (Fig. (Fig.9A).9A). Moreover, the polymerizing activity of the 17-mer BI-1 peptide is enhanced by increasing ionic strength, while the polymerization of actin by polycations is inhibited by increasing ionic strength, as shown in the case of the polycation spermine (data not shown).

BI-1 is a tamoxifen-regulated protein in tumor tissues from breast cancer patients (9). In another study, BI-1 prevented certain breast cancer cells from undergoing apoptosis, explaining the role of BI-1 in resistance to anticancer therapy (12). Studies to better understand the molecular basis of carcinogenesis and elucidate its signal transduction pathways have recently focused on BI-1 and its roles in the process of cancer development. In this field, the molecular mechanisms of BI-1 with respect to cancer-related characteristics, including cell adhesion, need to be more clearly studied in the near future.


This study was supported by a grant (no. A08144) from the Korea Healthcare Technology R&D Project, Ministry of Health, Welfare, and Family Affairs, South Korea; by a Research Program for New Drug Target Discovery grant from the Ministry of Education, Science & Technology; and partly by Korea Science and Engineering Foundation (R01-2007-000-20275-0), South Korea.


[down-pointing small open triangle]Published ahead of print on 1 February 2010.


1. Anand-Apte, B., and B. Zetter. 1997. Signaling mechanisms in growth factor-stimulated cell motility. Stem Cells 15:259-267. [PubMed]
2. Arredouani, A., Y. Guiot, J. C. Jonas, L. H. Liu, M. Nenquin, J. A. Pertusa, J. Rahier, J. F. Rolland, G. E. Shull, M. Stevens, F. Wuytack, J. C. Henquin, and P. Gilon. 2002. SERCA3 ablation does not impair insulin secretion but suggests distinct roles of different sarcoendoplasmic reticulum Ca2+ pumps for Ca2+ homeostasis in pancreatic beta-cells. Diabetes 51:3245-3253. [PubMed]
3. Bailly, M. 2003. Connecting cell adhesion to the actin polymerization machinery: vinculin as the missing link? Trends Cell Biol. 13:163-165. [PubMed]
4. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. [PubMed]
5. Reference deleted.
6. Camello, C., R. Lomax, O. H. Petersen, and A. V. Tepikin. 2002. Calcium leak from intracellular stores—the enigma of calcium signalling. Cell Calcium 32:355-361. [PubMed]
7. Chernyavsky, A. I., J. Arredondo, E. Karlsson, I. Wessler, and S. A. Grando. 2005. The Ras/Raf-1/MEK1/ERK signaling pathway coupled to integrin expression mediates cholinergic regulation of keratinocyte directional migration. J. Biol. Chem. 280:39220-39228. [PubMed]
8. Cristofanilli, M., and A. Akopian. 2006. Calcium channel and glutamate receptor activities regulate actin organization in salamander retinal neurons. J. Physiol. 575:543-554. [PubMed]
9. del Carmen Garcia Molina Wolgien, M., I. D. da Silva, F. E. Villanova, A. Yumi Otsuka, R. C. Borra, L. F. Lima Reis, A. F. Carvalho, E. C. Baracat, and L. H. Gebrim. 2005. Differential gene expression assessed by cDNA microarray analysis in breast cancer tissue under tamoxifen treatment. Eur. J. Gynaecol. Oncol. 26:501-504. [PubMed]
10. Dominguez, R. 2004. Actin-binding proteins—a unifying hypothesis. Trends Biochem. Sci. 29:572-578. [PubMed]
11. Gilon, P., A. Arredouani, P. Gailly, J. Gromada, and J. C. Henquin. 1999. Uptake and release of Ca2+ by the endoplasmic reticulum contribute to the oscillations of the cytosolic Ca2+ concentration triggered by Ca2+ influx in the electrically excitable pancreatic B-cell. J. Biol. Chem. 274:20197-20205. [PubMed]
12. Grzmil, M., S. Kaulfuss, P. Thelen, B. Hemmerlein, S. Schweyer, S. Obenauer, T. W. Kang, and P. Burfeind. 2006. Expression and functional analysis of Bax inhibitor-1 in human breast cancer cells. J. Pathol. 208:340-349. [PubMed]
13. Hage, B., K. Meinel, I. Baum, K. Giehl, and A. Menke. 2009. Rac1 activation inhibits E-cadherin-mediated adherens junctions via binding to IQGAP1 in pancreatic carcinoma cells. Cell Commun. Signal. 7:23. [PMC free article] [PubMed]
14. Harima, Y., A. Togashi, K. Horikoshi, M. Imamura, M. Sougawa, S. Sawada, T. Tsunoda, Y. Nakamura, and T. Katagiri. 2004. Prediction of outcome of advanced cervical cancer to thermoradiotherapy according to expression profiles of 35 genes selected by cDNA microarray analysis. Int. J. Radiat. Oncol. Biol. Phys. 60:237-248. [PubMed]
15. Hehlgans, S., M. Haase, and N. Cordes. 2007. Signalling via integrins: implications for cell survival and anticancer strategies. Biochim. Biophys. Acta 1775:163-180. [PubMed]
16. Hood, J. D., and D. A. Cheresh. 2002. Role of integrins in cell invasion and migration. Nat. Rev. Cancer 2:91-100. [PubMed]
17. Igney, F. H., and P. H. Krammer. 2002. Death and anti-death: tumour resistance to apoptosis. Nat. Rev. Cancer 2:277-288. [PubMed]
18. Johnson, B. D., and L. Byerly. 1993. A cytoskeletal mechanism for Ca2+ channel metabolic dependence and inactivation by intracellular Ca2+. Neuron 10:797-804. [PubMed]
19. Kim, H. R., G. H. Lee, K. C. Ha, T. Ahn, J. Y. Moon, B. J. Lee, S. G. Cho, S. Kim, Y. R. Seo, Y. J. Shin, S. W. Chae, J. C. Reed, and H. J. Chae. 2008. Bax inhibitor-1 is a pH-dependent regulator of Ca2+ channel activity in the endoplasmic reticulum. J. Biol. Chem. 283:15946-15955. [PubMed]
20. Kirfel, G., A. Rigort, B. Borm, and N. Herzog. 2004. Cell migration: mechanisms of rear detachment and the formation of migration tracks. Eur. J. Cell Biol. 83:717-724. [PubMed]
21. Kouyama, T., and K. Mihashi. 1981. Fluorimetry study of N-(1-pyrenyl)iodoacetamide-labelled F-actin. Local structural change of actin protomer both on polymerization and on binding of heavy meromyosin. Eur. J. Biochem. 114:33-38. [PubMed]
22. Kruchten, A. E., E. W. Krueger, Y. Wang, and M. A. McNiven. 2008. Distinct phospho-forms of cortactin differentially regulate actin polymerization and focal adhesions. Am. J. Physiol. Cell Physiol. 295:C1113-C1122. [PubMed]
23. Lau, P. M., R. S. Zucker, and D. Bentley. 1999. Induction of filopodia by direct local elevation of intracellular calcium ion concentration. J. Cell Biol. 145:1265-1275. [PMC free article] [PubMed]
23a. Lee, G. H., C. Yan, S. J. Shin, S. C. Hong, T. Ahn, A. Moon, S. J. Park, Y. C. Lee, W. H. Yoo, H. T. Kim, D. S. Kim, S. W. Chae, H. R. Kim, and H. J. Chae. Bax inhibitor-1 enhances cancer metastasis by altering glucose metabolism and activating the sodium-hydrogen exchanger: the alteration of mitochondrial function. Oncogene, in press. [PubMed]
24. Machesky, L. M., and A. J. Hall. 1997. Role of actin polymerization and adhesion to extracellular matrix in Rac- and Rho-induced cytoskeletal reorganization. J. Cell Biol. 138:913-926. [PMC free article] [PubMed]
25. McGuffin, L. J., K. Bryson, and D. T. Jones. 2000. The PSIPRED protein structure prediction server. Bioinformatics 15:404-405. [PubMed]
26. McNiven, M. A., M. Baldassarre, and R. Buccione. 2004. The role of dynamin in the assembly and function of podosomes and invadopodia. Front. Biosci. 9:1944-1953. [PubMed]
27. Momboisse, F., S. Ory, V. Calco, M. Malacombe, M. F. Bader, and S. Gasman. 2009. Calcium-regulated exocytosis in neuroendocrine cells: intersectin-1L stimulates actin polymerization and exocytosis by activating Cdc42. Ann. N. Y. Acad. Sci. 1152:209-214. [PubMed]
28. Nobes, C. D., and A. Hall. 1995. Rho, Rac, and Cdc42 regulate the assembly of multi-molecular focal complexes associated with actin stress fibers, lamellipodia and filopodia. Cell 81:53-62. [PubMed]
29. Peacock, J. G., A. L. Miller, W. D. Bradley, O. C. Rodriguez, D. J. Webb, and A. J. Koleske. 2007. The Abl-related gene tyrosine kinase acts through p190RhoGAP to inhibit actomyosin contractility and regulate focal adhesion dynamics upon adhesion to fibronectin. Mol. Biol. Cell 18:3860-3872. [PMC free article] [PubMed]
30. Pertz, O., L. Hodgson, R. L. Klemke, and K. M. Hahn. 2006. Spatiotemporal dynamics of RhoA activity in migrating cells. Nature 440:1069-1072. [PubMed]
31. Schaller, M. D., J. D. Hildebrand, and I. T. Parsons. 1999. Complex formation with focal adhesion kinase: a mechanism to regulate activity and subcellular localization of Src kinases. Mol. Biol. Cell 10:3489-3505. [PMC free article] [PubMed]
32. Spudich, J. A., and S. Watt. 1971. The regulation of rabbit skeletal muscle contraction. I. Biochemical studies of the interaction of the tropomyosin-troponin complex with actin and the proteolytic fragments of myosin. J. Biol. Chem. 246:4866-4871. [PubMed]
33. Tengholm, A., B. Hellman, and E. Gylfe. 2001. The endoplasmic reticulum is a glucose-modulated high-affinity sink for Ca2+ in mouse pancreatic beta-cells. J. Physiol. (Lond.) 530:533-540. [PubMed]
34. Tengholm, A., B. Hellman, and E. Gylfe. 2000. Mobilization of Ca2+ stores in individual pancreatic beta-cells permeabilized or not with digitonin or alpha-toxin. Cell Calcium 27:43-51. [PubMed]
35. Ueda, H., S. Saga, H. Shinohara, R. Morishita, K. Kato, and T. Asano. 1997. Association of the γ12 subunit of G proteins with actin filaments. J. Cell Sci. 110:1503-1511. [PubMed]
36. Vasioukhin, V., C. Bauer, M. Yin, and E. Fuchs. 2002. Directed actin polymerization is the driving force for epithelial cell-cell adhesion. Cell 100:209-219. [PubMed]
37. Worley, J. F., III., M. S. McIntyre, B. Spencer, R. Z. Mertz, M. W. Roe, and I. D. Dukes. 1994. Endoplasmic reticulum calcium store regulates membrane potential in mouse islet beta-cells. J. Biol. Chem. 269:14359-14362. [PubMed]
38. Wu, R. W., F. S. Wang, J. Y. Ko, C. J. Wang, and S. L. Wu. 2008. Comparative serum proteome expression of osteonecrosis of the femoral head in adults. Bone 43:561-566. [PubMed]
39. Xu, Q., and J. C. Reed. 1998. Bax inhibitor-1, a mammalian apoptosis suppressor identified by functional screening in yeast. Mol. Cell 1:337-346. [PubMed]
40. Reference deleted.

Articles from Molecular and Cellular Biology are provided here courtesy of American Society for Microbiology (ASM)