Pancreatic-Derived Factor (FAM3B), a Novel Islet Cytokine, Induces Apoptosis of Insulin-Secreting β-Cells

PANDER, a novel cytokine identified and characterized by GlaxoSmithKline (King of Prussia, PA), belongs to a new cytokine family and encodes a 235-amino acid protein with a secretion signal peptide (GenBank accession no. AF494379). Our previous study has shown that PANDER is mainly expressed in the pancreas (17). The expression and purification of recombinant murine and human PANDER was performed as previously described (17). Murine PANDER was used in most of the experiments, except for the studies with human islets, which used human PANDER. All other materials were from Sigma unless otherwise indicated.

β-TC3 insulinoma cells were cultured in T175 cm flasks in complete RPMI-1640 (11 mmol/l glucose) supplemented with 10% fetal bovine serum (FBS), 100 units/ml penicillin, 100 μg/ml streptomycin, and 2 mmol/l l-glutamine at 37°C under conditions of 95% air and 5% CO₂. The medium was changed twice a week and on the day before the experiment. Cells were trypsinized weekly and were used exclusively between passages 40 and 53. For PANDER studies, β-TC3 cells were fed for 24 h before use and washed with prewarmed RPMI-1640 (3 mmol/l glucose, 1% FBS) twice. Cells were then cultured in RPMI-1640 (1% FBS, 3 mmol/l glucose) supplemented with PANDER as indicated. Vehicle-treated cells served as controls.

α-TC3 cells were cultured in T175 cm flasks in complete Dulbecco’s modified Eagle’s medium (DMEM) (25 mmol/l glucose) supplemented with 15% FBS, 100 units/ml penicillin, 100 μg/ml streptomycin, and 2 mmol/l l-glutamine at 37°C under conditions of 95% air and 5% CO₂. The medium was changed twice a week and on the day before the experiment. Cells were trypsinized weekly and were used exclusively between passages 37 and 41. For PANDER studies, α-TC3 cells were fed for 24 h before use and washed with prewarmed DMEM (25 mmol/l glucose, 5% FBS) twice. Cells were then cultured in DMEM (1% FBS, 25 mmol/l glucose) supplemented with PANDER as indicated.

Male Sprague-Dawley rats or C57 black mice (Charles River Laboratories, Boston, MA) were injected with Nembutal. After the rats were anesthetized, the bile duct was cannulated, and the pancreas was inflated with ∼20 ml Hank’s balanced buffer. The inflated pancreas was removed and cleaned of its lymph nodes, fat, blood vessels, and bile duct. Tissue was digested with collagenase P (Roche Molecular Biochemicals) as previously described (18,19) and purified on a discontinuous Ficoll gradient. Isolated islets were washed and cultured in complete CMRL-1066 (supplemented with 10% FBS, 5.6 mmol/l glucose, 2 mmol/l l-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin) culture medium at 37°C (95% air and 5% CO₂). Human islets were obtained from the Islet Isolation Core of the Juvenile Diabetes Foundation-Penn Islet Transplantation Center. Human islets were cultured in CMRL (lot no. 99603007, supplemented with 10% FBS, 2 mmol/l l-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin; Mediatech, Cellgro, VA) at 37°C under 95% air and 5% CO₂.

Human islets were centrifuged at 600g for 5 min, and the pellet of islets was embedded with frozen tissue embedding media (Fisher Diagnostic). The islet mass was then cut into sections of 4 μm and stored at −80°C until use. α-TC3 cells and β-TC3 cells were seeded and cultured in the chambers of a poly-l-lysine-coated Lab-Tek chamber slide system (Nalge Nunc). On the day of the experiment, sections were washed twice with cold PBS, fixed with acetone at −20°C for 5 min, and then incubated with 5% BSA in PBS for 30 min at room temperature. Sections were then treated with a mixture of antibodies of goat anti-insulin (1:50 dilution; Santa Cruz) and rabbit anti-PANDER (1:50 dilution; GlaxoSmithKline) or a mixture of goat anti-glucagon (1:50 dilution; Santa Cruz) and rabbit anti-PANDER (1:50 dilution; GlaxoSmithKline) for 2 h at room temperature. The sections were washed and treated with a mixture of donkey anti-goat fluorescein isothiocyanate (FITC; 1:100 dilution) and donkey anti-rabbit rhodamine (1:200 dilution; Jackson ImmunoResearch, Chester, PA) in 2% normal donkey serum in PBS for 1 h in the dark. The sections were washed with ice-cold PBS again and mounted with PermaFluor reagent (Immunon). Samples were visualized by fluorescent confocal microscopy (Leica). Images were captured using Leica TCS SP2 fluorescent microscopy using a HeNe-argon laser as a light source. Leica confocal software was used to operate the microscopy and laser source. Each image was collected with excitation at 488 nm, emission at 520 nm (green, FITC), and with excitation at 543 nm, emission at 610 nm (red, rhodamine), concurrently. The images were analyzed using Leica confocal software.

The C,N-diphenyl-N′4–5-dimethyl thiazol-2-yl tetrazolium bromide (MTT) assay is an indirect measure of β-cell viability (20). The assay is based on the ability of viable cells to reduce MTT to insoluble colored formazan crystals. β-TC3 cells were plated in 24-well plates (1 × 10⁶) on the day before the experiment. After culture with or without PANDER, cells were washed twice with Krebs-HEPES buffer (115 mmol/l NaCl, 24 mmol/l NaHCO₃, 5 mmol/l KCl, 1 mmol/l MgCl₂, 2.5 mmol/l CaCl₂, 25 mmol/l HEPES, and 1% BSA, pH 7.4) with no glucose and incubated in 1 ml Krebs-HEPES buffer (0 mmol/l glucose) with 0.5 mg/ml MTT for 60 min at 37°C. The supernatant was discarded, and cells in each well were lysed with 500 μl of 2-propanol and incubated for 1 h at room temperature. The optical density of the resultant colored 2-propanol was measured at 560 nm on a MicroKinetics plate reader.

Annexin V is a Ca²⁺-dependent phospholipid-binding protein with a high affinity for phosphatidylserine (PS). In normal cells, PS is located on the cytoplasmic surface of the cell membrane. In apoptotic cells, PS is translocated from the inner to the outer surface of the cell membrane. Annexin V labeled with a fluorophore can identify apoptotic cells by binding to PS exposed on the outer leaflet. β-TC3 cells were treated with vehicle or with 4 nmol/l PANDER for 2 days and then washed once with ice-cold PBS followed by annexin V binding buffer (10 mmol/l HEPES, 140 mmol/l NaCl, and 2.5 mmol/l CaCl₂, pH 7.4). Cells were then resuspended with annexin V binding buffer and incubated with 25 μl annexin V/100 μl binding buffer and 10 μg/ml propidium iodide (PI) for 15 min at room temperature in the dark. Cells were washed again with annexin V binding buffer and analyzed by confocal microscopy with excitation at 488 nm (green, annexin V) and 543 nm (red, PI), concurrently.

PI is a highly polar dye that penetrates cells with damaged membranes and stains the nuclei red. After β-TC3 cells were cultured with PANDER, PI solution was added to each well (final 10 μg/ml) and incubated for 15 min at room temperature in the dark. Islets were stained with 15 μg/ml PI for 45 min at room temperature. Cell images were visualized by a Nikon fluorescence microscope (Diaphot) connected with a Hamamatsu digital camera (C4742-95) at 40× magnification and were analyzed with SimplePCI software (Compix Imaging Systems, Cramberry, PA) with excitation at 488 nm and emission at >510 nm. For each well, three or more random pictures (at 40×) were acquired under bright and fluorescence microscopy. Apoptotic cells were detected by their fragmented red nuclei, whereas necrotic cells were detected by their intact red nuclei. The percentage of apoptotic and necrotic cells was determined by counting all of the red-stained cells divided by all of the cells in the bright-field pictures. A minimum of 500 cells was counted per well. For quantitative analysis of islet viability, the PI fluorescence intensity of the total islets in each condition was measured with a fluorometer (Victor 2000; PerkinElmer).

An in situ cell death detection kit (Boehringer Mannhein, Indianapolis, IN) was used to quantify apoptotic cells. In this method, terminal deoxynucleotidyl transferase was used to catalyze the polymerization of fluorescein-labeled nucleotides to free 3-OH termini of DNA strand breaks. After 2 days culture with vehicle or with 4 nmol/l PANDER, β-TC3 cells were trypsinized and washed twice with cold PBS/1% BSA. Cells were then fixed with 200 μl 4% paraformaldehyde and incubated 30 min at room temperature. After rinsing with PBS, cells were resuspended with 250 μl of permeabilization solution (0.1% Triton X-100 in 0.1% sodium citrate) and incubated on ice for 10 min. Then, 50 μl of terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) reaction mixture was added to samples and the positive controls (50 μl of label solution only was added to the negative controls), and cells were incubated at 37°C for 1 h. Apoptotic cells were identified by FITC staining and analyzed by flow cytometry. A Coulter EPICS Elite flow cytometer (Beckman-Coulter, Hialeah, FL) equipped with a 5-W argon laser operated at 488 nm and 260 mW output was used for all studies. Monomeric forms of the β-TC3 cells were electronically gated based on forward- and side-scatter measurements to exclude cell aggregates from evaluation. Fluorescence signals were collected with a photomultiplier tube configured with 550 nm dichroic and 525 nm band pass filters. A total of 10,000 events were collected into a 4-decalog single-parameter histogram for each sample. Percent positive cells were determined based on the evaluation of cells treated with TUNEL reagents lacking TdT (terminal deoxynucleotidyl transferase) using a cursor setting that yielded <2% positive cells.

Mouse islets were cultured in complete CMRL-1066 in 6-cm dishes with vehicle or with 4 nmol/l PANDER for 4–5 days. Islet preparations were then washed briefly with prewarmed serum-free medium and fixed with 2.5% glutaraldehyde and 2% paraformaldehyde in sodium cacodylate buffer for 4 h. The islets were collected from the dish to make a pellet. After washing, osmication, and dehydration with ethanol and propylene oxide, the pellets were embedded in EM-Bed 812 and polymerized at 70°C for 48 h. Semithin sections (1 μm) were stained with Toluidine blue to screen general cell morphology. Ultrathin sections (80 nm) were cut with a diamond knife, stained with uranyl acetate and lead citrate, and examined with a JEOL-1010 TEM (transmission electron microscopy) operated at the accelerated voltage of 80 kV.

A Cayman chemical nitrate/nitrite assay kit was used to detect the sum of both nitrite (NO₂⁻) and nitrate (NO₃⁻). Two hundred islets were transferred to each well in a 96-well plate with 200 μl of complete CMRL, and they were incubated with the reagents indicated for 2 days. Then, 100 μl of cell-free culture media was mixed with 100 μl of Greiss reagent and incubated for 15 min, and the absorbance was determined at 540 nm with a Microkinetics plate reader (Fisher) (21).

β-TC3 cells were plated on microscope glass coverslips in 6-well plates 2 days before the experiment and then cultured in RPMI-1640 supplemented with 1% FBS and 3 mmol/l glucose with vehicle or with 4 nmol/l PANDER. On the day of the experiment, cells were loaded with 2.5 μmol/l fura-2 (Molecular Probes, Eugene, OR) at 37°C and 5% CO₂ for 30 min. The coverslip with the cells was then mounted in a perfusion chamber placed on a homeothermic platform set at 37°C and visualized by a Nikon fluorescence microscope (Diaphot) connected with a Hamamatsu digital camera (C4742-95), at 20× magnification. The cells were superfused with Krebs-HEPES buffer at 37°C. Fura-2 was excited at 340 nm and 380 nm and emitted at 510 nm. The fluorescence images were analyzed by SimplePCI software (Compix Imaging Systems). The concentration of Ca²⁺ was calculated by comparing the ratio of fluorescence at each pixel to an in vitro two-point calibration curve. The Ca²⁺ concentration is presented by averaging the values of all pixels of a single cell.

β-TC3 cells were cultured in 10-cm dishes in RPMI-1640 supplemented with 1% FBS, 3 mmol/l glucose, 100 units/ml penicillin, 100 μg/ml streptomycin, and 2 mmol/l l-glutamine at 37°C under conditions of 95% air and 5% CO₂. Cells were treated with vehicle or with 4 nmol/l PANDER for 20 h. Cell lysates were prepared and extracted with SDS sample buffer. Samples were analyzed by 15% SDS-PAGE and electrotransferred to Hybond nitrocellulose membranes, which were then probed with antibodies to cleaved caspase-3 (Cell Signal) and PANDER (GlaxoSmithKline). For PANDER expression, only the control lysates were used. Immunoblot analysis was performed using enhanced chemiluminescence Western blotting detection reagents (Amersham Phamacia Biotech).

Data are presented as the means ± SE. Statistical significance of differences between groups was analyzed by unpaired Student’s t test or by one-way ANOVA when more than two groups were compared.

PANDER was clearly localized to the islets of Langerhans of human pancreas and colocalized with insulin as well as glucagon (Fig. 1), suggesting that PANDER is expressed in pancreatic β- and α-cells in the human islets. Immunofluorescent double-staining of insulin and PANDER in β- and α-TC3 cells (Fig. 1) also demonstrated the colocalization. A protein band around 23 kDa, the expected mass of PANDER, was detected in Western blots of lysates from insulin-secreting cell lines (β-TC3) and blotted with anti-PANDER antibody (Fig. 9A).

The effect of PANDER on rat and human islet-cell viability was measured by confocal microscopy analysis and quantitative assay of fluorescence intensity after PI staining. After 7–9 days culture with 4 nmol/l PANDER, there was an increase in dead cells in the PANDER-treated rat and human islets (Figs. 2B and 3B) compared with vehicle-treated islets (Figs. 2A and 3A). As shown in Figs. 2C and 3C, the increase in PI-staining cells was significantly higher in PANDER-treated rat and human islets compared with control. Electron microscopy analysis of PANDER-treated mouse islets showed that PANDER induced different stages of apoptotic morphologic changes in the cells, which included condensation in the nuclei, granule dilution, clumped and condensed mitochondria, and nuclear envelopes with irregular dilation in the early stage (Fig. 4B). PANDER also induced flocculent density of mitochondria, more condensation in the nuclei, and more diluted granules in the late stage (Fig. 4C).

To determine the dose-dependency of PANDER’s cytotoxic effects, insulin-secreting β-TC3 cells were incubated in RPMI-1640 with increasing amounts of PANDER for 48 h and then evaluated for cell viability with the MTT assay. As shown in Fig. 5A, a low dose of 4 pmol/l PANDER had no effect on β-TC3 cell viability, but a dose-dependent inhibitory effect on cell viability was observed with PANDER concentrations >40 pmol/l. The optical densities in the MTT assay were 0.70 ± 0.05 (vs. control 0.77 ± 0.05, P > 0.05), 0.61 ± 0.05 (vs. control P < 0.05), 0.54 ± 0.05 (vs. control P < 0.01), and 0.48 ± 0.05 (vs. control P < 0.001) for 4, 40, and 400 pmol/l and 4 nmol/l of PANDER, respectively. Furthermore, exposure of β-TC3 cells to 4 nmol/l PANDER for 24 h to 5 days induced a time-dependent inhibitory effect on β-TC3 cell viability. Cell viability was inhibited by 27, 43, 55, and 75% at 2, 3, 4, and 5 days, respectively, even though PANDER had no significant effect at 24 h (Fig. 5B).

To determine the cytotoxic effect of PANDER on glucagon-secreting α-cells, α-TC3 cells were treated with or without 40 or 400 pmol/l PANDER for 48 h before cell viability was measured by the MTT method. Cell viability was decreased to 82.8 ± 3.3 or 67.0 ± 3.5% compared with control by 40 or 400 pmol/l PANDER, respectively (P < 0.05) (Fig. 5C). These results imply that PANDER is also cytotoxic to α-cells.

Both annexin V fluorescence staining and PI fluorescence staining were used to analyze the morphologic changes of β-TC3 cells treated by PANDER. After a 48-h culture with 4 nmol/l PANDER or with vehicle (control), PANDER-treated cells had a much higher percentage of annexin V-positive apoptotic cells (green membrane fluorescence) and PI-positive necrotic cells (red nuclei fluorescence) than vehicle-treated cells, confirming the findings of the MTT assay (Fig. 6). To obtain a quantitative assessment, PANDER-treated β-TC3 cells were stained with PI and counted. Compared with control, there was an almost threefold increase of necrotic cells in the PANDER-treated cells than the control cells. The presence of different concentrations of FBS (1–10%) did not prevent PANDER-induced toxicity (Fig. 7). Several of the PANDER-treated β-TC3 cells showed condensation and fragmentation of the nuclei, suggesting that PANDER induced apoptosis in these cells (Fig. 7).

To further analyze PANDER-induced apoptosis in β-TC3 cells, DNA fragmentation was measured by TUNEL labeling of β-TC3 cells. The TUNEL method was first validated using negative and positive controls. First, the number of positive cells increased to 94.1 ± 1.3% after DNase treatment of β-cell DNA compared with 1.3 ± 0.2% in the negative control. Second, after H₂O₂ killing of β-cells, the number of positive cells increased to 65.9 ± 18.2%. As shown in Fig. 8, PANDER (48-h treatment) caused a significant right shift of TUNEL labeling of β-TC3 cells analyzed by flow cytometry, resulting in a highly significant 7.5-fold increase in apoptotic β-TC3 cells compared with vehicle (16.4 ± 5.1 vs. 2.2 ± 1.0, n = 7, P < 0.05).

The level of cleaved caspase-3 was significantly increased in cells treated with PANDER for 20 h compared with control cells, as assessed by immunoblotting. Densitometric analysis of the cleaved caspase-3 bands showed an increase to 1,029 ± 34% vs. control 100 ± 16% (P < 0.0001), with no changes in caspase-3 between the two groups (Fig. 9B). Carbachol (1 mmol/l) induced a rapid and transient increase of cytosolic intracellular Ca²⁺ concentration ([Ca²⁺]_i) in both vehicle- and PANDER-treated β-TC3 cells. The peak was about twofold higher than the average baseline (230.0 ± 12.6 vs. 112.2 ± 2.7 nmol/l Ca²⁺ in the control cells, 201.3 ± 7.3 vs. 114.9 ± 3.8 nmol/l Ca²⁺ in the PANDER-treated cells; n = 7). There was no significant difference between the control and the PANDER-treated cells in both the basal condition as well as the carbachol-stimulated condition. Also, potassium induced a rapid and fourfold increase of cytosolic [Ca²⁺]_i in both vehicle- and PANDER-treated β-TC3 cells (467.5 ± 19.6 vs. 152.1 ± 5.3 nmol/l Ca²⁺ in the control cells, 422.5 ± 13.7 vs. 160.4 ± 6.3 nmol/l Ca²⁺ in the PANDER-treated cells; n = 7). There were also no significant differences between the control and the PANDER-treated cells under basal conditions and after potassium stimulation. To determine whether PANDER cytotoxicity was caused by elevated nitric oxide (NO) production, we measured levels of NO end products in the culture media. Mouse PANDER did not induce NO production at either of the concentrations tested (data not shown).

PANDER is a member of a new cytokine family identified by a search for novel proteins with the unique architecture and topology of the short-chain, four-helix bundle cytokines (17). Although its function is unknown, PANDER was found to be specifically expressed in pancreatic islets of Langerhans, including β-cells as well as α-cells, and is present in insulin-secreting β-cell lines. To elicit the potential role of PANDER in the pancreatic islets, we investigated whether PANDER affects pancreatic β-cell, α-cell, and islet viability. We show that treatment of β-cells, α-cells, and islets with PANDER causes cell death. PANDER inhibits β-cell viability in a dose-dependent as well as time-dependent manner. It induces β-cell apoptosis through caspase-3 activation.

β-Cell destruction in type 1 diabetes is the result of a chronic inflammatory lesion mediated by interaction between macrophages, CD4⁺ cells, and CD8⁺ T-cells. The role of the inflammatory cytokines interleukin-1β, tumor necrosis factor-α, and interferon-γ produced by activated macrophages and CD8⁺ T-cells in this process has been reported by many groups (7,22). Cytokines are important mediators of this process (9,15,23,24). There is extensive evidence that these cytokines cause β-cell destruction through necrosis or apoptosis (13), although apoptosis is considered the main mode of β-cell destruction in immune-mediated type 1 diabetes (10–12). However, the exact mechanisms underlying the apoptotic process are still controversial and obscure (25–29). The increased production of NO may be an important signal in cytokine-induced cell death (9,23). In rat islets, NO can cause β-cell toxicity via different mechanisms (13,30), although in human pancreatic islets, cytokine-induced β-cell death may be independent of NO generation (2,15). PANDER, however, had no effect on NO production, suggesting that its mechanism of action is different from that of cytokines such as interleukin-1β and tumor necrosis factor-α.

Intracellular Ca²⁺ may have a role in apoptosis of the pancreatic β-cell (31,32). Increases in intracellular Ca²⁺ can promote endonuclease activation, mitochondrial depolarization, and permeability transition and subsequent cytochrome c release, caspase activation, and apoptotic cell death (33,34). Other intracellular cysteine proteases, such as calpains, can be initiated by Ca²⁺ and are involved in apoptosis (32). It has also been reported that high glucose and tolbutamide can induce apoptosis in pancreatic β-cells in a Ca²⁺-dependent manner (35), and a calcium channel blocker can partially inhibit glibenclamide-induced β-cell apoptosis (33). In this study, we examined the effects of PANDER on Ca²⁺ homeostasis in β-cells. Neither basic [Ca²⁺]_i nor carbachol- or potassium-induced [Ca²⁺] increase was affected by PANDER. Carbachol and potassium utilize different pathways to stimulate cytosolic Ca²⁺. Carbachol releases Ca²⁺ from the endoplasmic reticulum, whereas potassium depolarizes the membrane, resulting in Ca²⁺ influx through voltage-dependent Ca²⁺ channels (36). Clearly, PANDER does not affect Ca²⁺ homeostasis in pancreatic β-cells. Therefore, its pro-apoptotic effects in pancreatic β-cells are Ca²⁺ independent.

Caspases are a family of proteases. They are present in most cells and are crucial mediators of apoptosis (37). Caspase-3 is at pivotal junction in the apoptotic pathway (38). It is activated by proteolytic cleavage into subunits of 19 and 17 kDa. Our data has shown that culture of β-ΤC3 cells with PANDER for 20 h significantly increased the 19- and 17-kDa cleaved caspase-3 subunit, suggesting that PANDER induces β-cell apoptosis via the caspase-3 pathway.

What is the role of PANDER in islet physiology and pathophysiology? PANDER is a novel, secreted cytokine-like protein present in β-cells of mouse, rat, and human pancreatic islets. Our initial observations indicate that recombinant PANDER has potent apoptotic effects on the β-cell that are analogous to those originally described for interleukin-1 (39,40). Although the receptor for PANDER has yet to be identified, the demonstration that it kills β-cells in vitro suggests that the β-cells could be a target for the action of PANDER. Clearly, an understanding of the mechanisms regulating PANDER production and secretion are crucial to understanding how a β-cell-specific secreted protein can induce apoptosis of islets. Whether PANDER is involved in the pathogenesis of diabetes remains to be determined and is currently under investigation.

This work was supported by grants from the Juvenile Diabetes Research Foundation and the National Institutes of Health (DK-49814) and a Juvenile Diabetes Research Foundation Postdoctoral Fellowship. The Radioimmunoassay Core and Biomedical Imaging Core of the Penn Diabetes Center are supported by National Institutes of Health Grant DK-19525.

We thank the following University of Pennsylvania Diabetes Center Core facilities: the Insulin RIA Core for analysis of insulin secretion samples and the Biomedical Imaging Core for use of the confocal microscope and electron microscope.