PANC-1 and Miapaca-2 cells from American Type Culture Collection (ATCC) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with glutamine containing 10% fetal bovine serum, penicillin, and streptomycin. Transfections were carried out with lipofectamine 2000 (Invitrogen) in OPTI-MEM (Invitrogen) according to the manufacturer’s protocol with cells plated 18 hr before transfection. The total amount of plasmid used per transfection was normalized with the corresponding empty vector constructs. For siRNA transfection, cells were transfected with AES specific siRNAs or control siRNAs (obtained from Ambion) with the Lipofectamine RNAiMAX reagent (Invitrogen) according to the manufacturer’s recommendation [11]. To generate stable Bit1 knockdown and control pools of clones, PANC-1 cells infected with controlor Bit1-shRNA lentiviral particles (Santa Cruz Biotechnology) in a six-well plate format in the presence of polybrene (5 ug/ml) [11,12]. 24 h postinfection, cells were treated with 2 ug/ml puromycin to select for stable control and Bit1 knockdown clones. Several puromycin resistant control and Bit1 knockdown clones were harvested by ring-cloning, and the level of Bit1 knockdown was confirmed by immunoblotting against a specific Bit1 antibody. Two controlshRNA clones and three positive Bit1shRNA knockdown clones were pooled together to generate the controlshRNA pool and Bit1shRNA pool, respectively. The resulting controlshRNA and Bit1shRNA pools were then again subjected to immunoblotting using a specific antibody to Bit1 to confirm the downregulation of Bit1 expression.

Poly(2-hydroxyethyl methacrylate) (Polyhema) and the mouse monoclonal anti-β-actin antibody were obtained from Sigma Chemical Co. (St. Louis, Mo). The mouse monoclonal anti-FLAG, anti-AES, and anti-B-actin antibodies were purchased from Sigma (St. Louis, MO). The caspase inhibitor zVad-fmk and etoposide were purchased from Calbiochem (La Jolla, CA). The anti-c-myc antibody was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The specific antibody to Bit1 used for immunoblotting was generated as described previously [9], and the affinity purified rabbit anti-Bit1 antibody (HPA012897, Sigma) was used for immunohistochemistry studies [12]. The mammalian expression vector encoding for mitochondrial Bit1 was generated as described previously [9]. The DDK-TLE1 plasmids were obtained from Origene (Rockville, MD).

Apoptosis was assessed by determining the level of cytosolic nucleosomal fragments with the use of Cell Death Detection ELISA kit (Roche Molecular Biochemicals), according to the manufacturer’s instructions [11]. To assess for anoikis cell death, cells were plated onto Polyhema coated 96 well plates in complete growth medium containing 0.5% methylcellulose at a density of 1.0 × 10⁴/well at various time points as previously described [11,12]. Detached cells were then collected and subjected to the Cell Death ELISA apoptosis assay. In some experiments, cell viability was quantified by clonogenic assay [13] or with the use of MTT (Thiazolyl Blue Tetrazolium Bromide, Sigma). Briefly, fifteen µl of MTT solution (5 mg/ml in PBS) was added and further incubated for 4 h. The resulting MTT formazan was solubilized by addition of 100 µl of SDS solution (20% in 10 mM HCL), and the absorbance was measured 24 h at 550 nm and a reference wavelength of 690 nm using a microplate reader. The anchorage-independent growth of cells was measured using the 96-well plate format [11]. Briefly, 5000 cells in 0.3% agar solution was plated onto wells precoated with 0.6% agar in culture medium. The anchorage-independent growth of cells was quantified by alarmar blue staining (Invitrogen) and fluorescence reading at 530 - 560 nm excitation wavelength and 590 nm emission wavelength.

Protein preparation and western blotting were performed as described previously [11,12]. Briefly, cells were harvested 24 - 48 hr after transfection with various constructs or siRNAs by adding ice-cold NP-40 lysis buffer (1% NP-40; 20 mM Tris-HCL [pH 7.4]; 150 mM NaCl; 10% glycerol, 2 mM sodium vanadate; 1 mM henylmethylsulfonyl fluoride; 10 μg/ml leupeptine; and 5 μg/ml aprotinin) and incubating at 4˚C for 20 min. For immunoblot analysis, equal amounts of proteins were resolved on 4% - 20% gradient Tris-glycine gels (Invitrogen) and electrophoretically transferred to nitrocellulose membrane. The membranes were incubated with primary antibodies overnight at 4˚C followed by secondary antibodies conjugated with horseradish peroxidase. Membranes were developed using the ECL detection system. For co-immunoprecipitation assays, transfected cells were harvested in ice-cold Nonidet P-40 lysis buffer, and cell debris was removed by centrifugation. Flag-tagged Bit1 was immunoprecipitated with anti-Flag-agarose conjugate (Sigma) and thoroughly washed with lysis buffer. Bound proteins were resolved by SDS-PAGE, and western blotting was performed using anti-Flag or anti-Myc antibodies [10].

Pancreatic tumor tissue array slides were obtained from US Biomax, Inc. (Rockville, MD). The immunohistochemistry procedure was performed by Biomax Inc. on two tissue microarray slides. As described previously [11,12], tissue array slides were deparaffinised, hydrated and subjected to antigen retrieval. The slides were then incubated in 2.5% normal horse serum for 30 min at room temperature followed by incubation with the primary antibody (1:100 dilution) for 1 h at room temperature. Rabbit normal serum was used as negative control antibody to replace the primary antibody on control slide with 1 hr incubation. Tissue array slides were then washed and incubated with ImmPRESS reagent (Vector Laboratories) followed by treatment with peroxidise substrate DAB solution (DAKO Cytomation). Each of the experimental and control slide was scored for average staining intensity by two investigators with no knowledge of the pathologic status of the samples. These investigators scored staining intensity as 0, no staining; 1 low staining; 2 medium staining; or 3 high staining.

Data are presented as means (±S.E.). For western blots and anoikis assays, experiments were performed at least three times with duplicates. Statistical differences between groups were established at a P value < 0.05 using the two-tailed Student’s t test. For pancreatic tumor tissue array analysis, a one-way ANOVA with subsequent post hoc testing using the Tukey-Kramer multiple comparison test was used to compare the average staining intensity of each case type [11,12]. All calculations were done using the NCSS statistical software (NCSS, Kasville, UT).

Since anoikis resistance is a primary determinant of tumorigenesis and/or metastasis [14], we hypothesized that the anoikis effector Bit1 is downregulated or dysfunctional in advanced tumors. We performed immunohistochemistry to examine Bit expression in a human pancreatic tumor tissue microarray with the use of an affinity purified polyclonal antihuman Bit1 antibody (Sigma). This polyclonal anti-Bit1 antibody was developed and its specificity validated by the Human Proteome Resource (HPR) project [11,12]. Immunoreactivity was graded as 0 (none), 1 (slight), 2 (moderate), or 3 (strong). In normal pancreatic tissue, both the exocrine acinar cells and the islet cells showed strong cytoplasmic staining for Bit1 (Figure 1(a(i))). The strong Bit1 staining was particularly homogenous within the pancreatic acini. As compared to the strong immunoreactivity of Bit1 in the normal pancreatic tissue, a subset of advanced pancreatic tumors (20 of 42 (47%) of stage III (Figure 1(a(ii)) and (22 of 39 (56%) of stage IV (Figure 1(a(iii)) exhibited significantly reduced cytoplasmic Bit1 immunoreactivity. Quantification of the average Bit1 staining (Figure 1(b)) confirmed the reduction of Bit1 immunoreactivity of advanced pancreatic tumor tissues relative to its normal counterpart. These findings indicate that Bit1 expression is lost in advanced pancreatic cancer, suggesting that downregulation of Bit1 expression may be an important event in the progression of pancreatic cancer.

To test the functionality of the Bit1 apoptotic pathway in pancreatic cancer cells, we ectopically overexpressed Bit1 in the cytoplasm of human pancreatic cancer cell line Miapaca-2 via transfection with either a vector control or N-terminally myc-tagged Bit1 construct [10]. Placement of tag at the N-terminal region of the Bit1 protein results in its cytoplasmic localization [9]. The cytoplasmic Bit1 transfected cells showed a dramatic decrease in cell viability 48 hr post transfection, as assessed by MTT (Figure 2(a)) and clonogenic (Figure 2(b)) assays. The reduction in cell viability in cytoplasmic Bit1 transfected cells was associated with increased levels of DNA-histone fragments (Figure 2(c)), indicating the induction of apoptotic form of cell death by Bit1. Similar induction of apoptosis was also observed upon targeting Bit1 expression in the cytoplasm of PANC-1 cells (Figure 2(d)). Consistent with previous findings [9], the induction of Bit1-induced cell death was not abrogated by the pan-caspase inhibitor zVAD-fmk, indicating the caspase-independence mode of Bit1-mediated apoptosis (Figure 2(e)).

We evaluated the effects of overexpressing the mito-

chondrial localized Bit1 on the anoikis sensitivity of pancreatic cancer cells. Mitochondrial Bit1 was overexpressed in Miapaca-2 cells via transfection of C-terminally myc tagged Bit1 or vector construct [9,10]. Exogenous expression of mitochondrial Bit1 induced significantly higher level of apoptosis following loss of cell attachment as compared with control cells (Figure 3(a)). In contrast, both control and mitochondrial Bit1-transfected cells showed similar levels of basal apoptosis in attached conditions (Figure 3(a)). To complement the results from exogenous mitochondrial Bit1 studies, we also examined the effect of downregulating Bit1 expression in PANC-1 cells, which exhibit moderate levels of endogenous Bit1. To this end, we generated stable Bit1 knockdown clones by infecting PANC-1 cells with lentiviral expression vectors for Bit1 specific shRNAs or control shRNAs [12]. As shown in Figure 3(b), the Bit1 expression was reduced by 70% in a pool of Bit1 shRNA clones as compared with control shRNA pool of clones. The control shRNA and Bit1 shRNA pools exhibited

similar levels of basal apoptosis (Figure 3(c)) in monolayer culture. In contrast, when cultured without anchorage, the Bit1 shRNA pool displayed a significantly decreased apoptosis than the control shRNA pool. Taken together, these results indicate that Bit1 is an anoikis effector in pancreatic adenocarcinoma cells.

Since anoikis resistance is a requirement of anchorageindependent growth, a hallmark of transformation [14, 15], we then examined the effect of exogenous mitochondrial Bit1 expression on the growth of Miapaca-2 cells in a soft agar assay. Consistent with their enhanced anoikis sensitivity (Figure 3(a)), mitochondrial Bit1 transfected cells showed significant reduction in anchorage-independent growth relative to control transfected cells (Figure 4(a)). The Bit1 mitochondrial-transfected cells resulted in smaller and fewer colonies than the control cells (Figures 4(b) and (c)).

Since chemoresistance is an important determinant of pancreatic tumor aggressiveness, we examined whether induction of the Bit1 apoptotic pathway may enhance the chemosensitivity of the highly chemoresistant Miapaca-2 cells. We overexpressed the mitochondrial localized Bit1 in Miapaca-2 cells and examined their responsiveness to the chemotherapeutic agent, etoposide. While vector transfected cells showed no or little response to low concentrations of etoposide, the Bit1 transfected cells exhibited significant apoptosis at these conditions (Figure 4(d)). At much higher doses of etoposide, the induction of cell death became evident in control cells albeit at a significantly lower level relative to mitochondrial Bit1 transfected cells. These findings indicate that mitochondrial Bit1 enhances the sensitivity of Miapaca-2 cells to etoposide-induced apoptosis.

We have previously shown that Bit1, upon its release to or following its ectopic expression in the cytoplasm, associates with the groucho related AES protein to induce apoptosis [9,10]. To address the functional requirement of AES in Bit1 apoptosis function in pancreatic cancer cells, we evaluated the effects of alteration of AES expression on the sensitivity of cells to Bit1 apoptosis. Specific downregulation of AES expression in Miapaca- 2 cells attenuated cytoplasmic Bit1 apoptosis (Figures 5(a) and (b)). Conversely, exogenous AES expression further enhanced the induction of apoptosis by Bit1 (Figure 5(c)). We then evaluated the association of Bit1 and AES protein via co-immunoprecipitation studies. Miapaca-2 cells were transfected with N-terminally flagtagged Bit1 (cyto Bit1) and myc-tagged AES. Immunoprecipitation with anti-flag and blotting of the precipitate with anti-myc antibodies revealed Bit1/AES complexes

in the transfected cells (Figure 5(d)). The groucho related nuclear TLE1 corepressor blocks the Bit1-AES complex formation (Figure 5(d)) and abrogates Bit1 apoptosis (Figure 5(e)). Taken together, these findings indicate that Bit1 associates and cooperates with AES to induce apoptosis in pancreatic cancer cells.