|Home | About | Journals | Submit | Contact Us | Français|
The novel islet-specific protein PANcreatic DERived Factor (PANDER; FAM3B) has beenextensively characterized with respect to the β–cell, and these studies suggest a potential function for PANDER in the regulation of glucose homeostasis. Little is known regarding PANDER in pancreatic α–cells, which are critically involved in maintaining euglycemia. Here we present the first report elucidating the expression and regulation of PANDER within the α–cell. Pander mRNA and protein aredetected in α–cells, with primary localization to a glucagon-negative granular cytosolic compartment. PANDER secretion from α–cells is nutritionally and hormonally regulated by L-arginine and insulin, demonstrating similarities and differences with glucagon. Signaling via the insulin receptor (IR) through the PI3K and Akt/PKB node is required for insulin-stimulated PANDER release. The separate localization of PANDER and glucagon is consistent with their differential regulation, and the effect of insulin suggests a paracrine/endocrine effect on PANDER release. This provides further insight into the potential glucose-regulatory role of PANDER.
PANcreatic DERived factor (PANDER, FAM3B) is a 235 amino acid protein which is one of four members of a novel family of cytokine-like proteins identified by a computational genomic search as previously described (Zhu et al., 2002). PANDER shares approximately 31.6 – 53.3% primary sequence homology with the other members of the FAM3 family (FAM3A, C and D), but without extensive homology to any other known cytokines. PANDER is predominantly expressed in the pancreatic islets of Langerhans, with minimal expression in the small intestine and prostate (Zhu et al., 2002).
PANDER is robustly expressed in the β–cells of islets and various β–cell lines (Cao et al., 2003;Burkhardt et al., 2005;Yang et al., 2005). Localization and regulatory similarities exist between PANDER and insulin in β–cells, including moderate co-localization within insulin granules (Xu et al., 2005) and co-secretion in response to glucose challenge (Burkhardt et al., 2005). The pander promoter is positively regulated by glucose in β–cell lines and islets (Burkhardt et al., 2005), with dose-dependent induction of pander mRNA and protein expression (Wang et al., 2008;Yang et al., 2005). Our limited characterization of a recently created global PANDER knockout mouse model reveals aberrations in homeostatic mechanisms including impaired glucose tolerance, insulin secretion, and reduced hepatic glucose production (HGP) (Burkhardt et al., 2009). Both in-vitro and in-vivo data generated thus far suggests a potential role for PANDER in glucose homeostasis.
Much of the work by our group and others has focused on PANDER expression and regulation in β–cell lines and islets with primary focus on similarities to insulin (Burkhardt et al., 2005;Burkhardt et al., 2008;Yang et al., 2005;Xu et al., 2005;Wang et al., 2008). However, PANDER is also detected in both α–cell lines and primary α–cells, possibly at greater levels in α– than in β–cells (Burkhardt et al., 2005;Cao et al., 2003). As the second largest islet cell population, α–cells are critically involved in the maintenance of euglycemia, primarily via secretion of the peptide hormone glucagon (FOA et al., 1952;FOA et al., 1957). Glucagon is released mainly in response to hypoglycemia in the fasted state in vivo, but is also regulated in response to other nutritional, hormonal, and neuroendrocrine stimuli (Gromada et al., 2007). Acting chiefly at the liver, glucagon antagonizes the action of insulin, with hyperglucagonemia and elevated hepatic glucose production observed in the diabetic state (Baron et al., 1987;Reaven et al., 1987). This underscores the physiological relevance of α–cells, glucagon, and potentially PANDER in maintaining normoglycemia.
With little known regarding PANDER in α–cells, we present the first report demonstrating robust PANDER expression within the pancreatic α–cell, differential cytoplasmic localization with respect to glucagon, and regulation in response to L-arginine, insulin, and glucose challenge, with mechanistic insight into insulin-stimulated PANDER release.
α–TC1-6 (courtesy of Dr. Bruce Verchere, University of British Colombia, Canada), α–TC3, α–TC-Lenti-IR (courtest of Dr. Michael Wheeler, University of Toronto, Canada), and C2C12 (ATCC) cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) (Invitrogen, 12491-015) while βTC3 cells were maintained in RPMI-1640 medium (Invitrogen, 12633-012), both supplemented with 10% fetal bovine serum, 100 U/ml penicillin, 100 μg/ml streptomycin, and 2 mM L-glutamine at 37°C, 95% air, 5% CO2. Cells were passaged weekly and were used at passages 26-45.
Detection of endogenous and over-expressed PANDER by immunoblot and immunohistochemical labeling of PANDER in pancreatic sections was performed using rabbit polyclonal PANDER anti-serum (Alpha Diagnostics, San Antonio, TX). For immuno-fluorescent (IF) labeling, the following antibodies were used: anti-glucagon monoclonal (R&D Systems, MAB1249); anti-glucagon polyclonal (Millipore, 4031-01); anti-FLAG (M2) polyclonal (Cell Signaling Technology, 2368); anti-GM130 monoclonal (BD Transduction Labs, 612008); anti-PDI (Affinity Bioreagents, MA3-019); anti-LAMP-1 (Developmental Studies Hybridoma Bank, DSHB 1D4B), and anti-Chromogranin-A polyclonal (Novus Biologicals, NB110-2475). Secondary antibodies were either Alexa-Fluor dye conjugated [DAR-Alexa 488; DAM-Alexa 594; DAGP-Alexa 647], biotinylated [DAR-biotin, DAM-biotin], or Cy-conjugated [DAM-Cy3, DAR-Cy2]. Tertiary streptavidin-HRP or streptavidin-Alexa-dye conjugates were employed for sequential LSAB-IHC / IF. For assessment of insulin signaling, the following antibodies were employed: pAkt-Thr308 (Cell Signaling, 4056), pAkt-Ser473 (Cell Signaling, 4060), and Akt (Cell Signaling, 9272).
Adherent α–TC1-6 cells were transiently transfected with 4 μg PANDER plasmid using Lipofectamine 2000 (Invitrogen, 11668-019) in OPTI-MEM (Invitrogen, 31985-062) according to manufacturer instructions. α–TC1-6 cells in suspension were nucleofected with 2 μg pCMV-PANDER-3-FLAG using the nucleofector kit V (Amaxa Biosystems, VCA-1003) and program G-016 according to manufacturer instructions.
Adherent cell layers were rinsed with cold Dulbecco’s Phosphate Buffered Saline (DPBS) (Invitrogen, 14040-117) and lysed using Roth’s Lysis Buffer (50 mM HEPES, 150 mM NaCl, 1% Triton X-100, 1 mg/ml aprotinin, and 0.01 mg/ml leupeptin), or 1X Cell Lysis Buffer (Cell Signaling, 9803) supplemented with phosphatase inhibitor cocktail 2 (Sigma, P5726) and 1 mM PMSF. Lysates were stored at −20°C until SDS-PAGE analysis.
For immunoblots, equal quantities of total lysate protein were boiled for 5 minutes in 5X SDS-sample buffer with 15 mg/ml dithiothreitol (DTT) and resolved on 12% polyacrylamide gels. Protein transfer to nitrocellulose or PVDF membranes was carried out and blotted using antibodies listed above and detection performed using ECL Advance Western blotting kit (GE Healthcare, RPN-2135). Immunoblot visualization was perfomed using the FujiFilm LAS-3000 Imaging system. Densitometric analyses were performed using the FujiFilm Multigauge software package.
Murine pancreatic islets were isolated from 6-10 week old C57BL/6 mice as previously described (Gao et al., 1999). Islets were hand-picked, counted, and dispersed using 1.8 U/ml dispase in Hank’s Balanced Salt Solution (HBSS) and gentle trituration. The resulting single-cell suspension was then sorted by fluorescence-activated cell sorting (FACS) as outlined previously (Pipeleers et al., 1985a).
Adherent cell layers were rinsed in cold DPBS, lysed, and RNA isolated using the RNeasy kit (QIAGEN, 74104) according to manufacturer instructions. DNAse treatment of all samples was performed. Equivalent input RNA was used for quantitative Taqman® RT-PCR. For insulin, glucagon, and β–actin amplification, commercially available Gene Expression Assays were used (Applied Biosystems), while for pander, either commercially available or the primer /probes listed below were employed. PANDER forward – 5′-TGCTCGCGGAGCTCATTC 3′, PANDER reverse – 5′-CCAATGCTTCGGATGTTGTAGA-3′, fluorescent (FAM) PANDER probe – 5′-TGACGTGCCCCTGTCCAGCACT-3′.
Primary α, β and non–α/non–β cell enriched populations obtained by FACS sorting of murine islets (as outlined above) were lysed and RNA isolated using the Taqman® PreAmp Cells-to-CT kit (Ambion, 4387299). 500 ng of input RNA was reverse transcribed and pre-amplified for target transcripts. Pre-amplification products were subsequently used in Taqman® qPCR.
Pancreata isolated from male C57BL/6 mice were fixed overnight in 1:10 buffered formalin (Fisher Scientific, SF93-4). Tissue was then processed, paraffin embedded, and 4 μm sections mounted onto ProbeOn™ Plus slides (FisherBiotech, 15-188-52) and slides stored at room temperature until stained. The sections were labeled for PANDER and glucagon using sequential LSAB-DAB and LSAB Immuno-fluorescent staining respectively. Slides were visualized using a confocal microscope (Leica, DM IRE2) coupled to a spectral confocal system (Leica, TCS SP2) and micrographs captured.
The coding sequence of the pander gene was cloned upstream (5′) of the triple-FLAG repeat moiety of the pCMV-3FLAG-3a plasmid (Stratagene, 240197), employing EcoRV and XhoI restriction endonuclease sites within the multiple cloning region and construct integrity was confirmed by restriction digest and sequence analysis.
α–TC1-6 cells nucleofected with pCMV-PANDER-3FLAG (described above) were cultured on glass coverslips (Fisherbrand®, 12-545-82-12CIR-1D) in complete DMEM. 48 hrs post-nucleofection, cells were fixed in 4% paraformaldehyde, then blocked using PSG: a solution of 0.01% saponin, 0.25% fish skin gelatin and 0.02% NaN3 in 1X PBS. Cells were then incubated in primary antibodies specific for FLAG and various sub-cellular markers using antibodies diluted in PSG for 1 hr at room temperature in a humidified environment. Coverslips were then rinsed three times with PSG at room temperature. Secondary antibodies in PSG were added to coverslips for 1 hr and cells again rinsed with PSG. Cells were then counterstained with DAPI and coverslips mounted using MOWIOL and stored at 4°C in the dark until visualized, and confocal micrographs captured.
α–TC1-6 and α–TC6-Lenti-IR cells were transiently transfected with pShuttle-PANDER (as previously outlined). 24 hrs post-transfection, cells were rinsed and pre-incubated in DMEM, 3 mM glucose, 1% FBS for 1 hr prior to stimulation with L-Arginine (Sigma, A5006), Insulin (Sigma, I9278), or a combination of both at the indicated concentrations. For insulin stimulation experiments, cells were serum starved at 1% FBS for at least 5 hrs prior to stimulation. Where indicated, 200 nM wortmannin (Sigma, W1628) in DMSO was added to pre-incubation media prior to insulin stimulation. Post-stimulation, media was collected and cells lysed. Harvested media was concentrated by ammonium sulfate protein precipitation (0.55 g/ml media). Equal quantities of media or lysate protein were resolved by SDS-PAGE and probed for PANDER. Densitometric analysis was carried out as previously described and relative secreted PANDER determined as ratio of PANDER in media to that in lysate for each sample, and expressed as fold over negative control.
Glucagon secreted into culture media from α–TC1-6 cells was determined using Glucagon EIA (Alpco, 48-GLUHU-E01) or RadioImmunoAssay (Linco, GL-32K) kits according to manufacturer instructions and normalized to total protein content of cell lysates.
Statistical analyses were performed using the GraphPad Prism 5.0 software package. Statistical significance of differences was determined by Student’s t-test or analysis of variance (ANOVA) where more than two groups were compared. Threshold of statistical significance defined as P<0.05.
To investigate the expression of PANDER in pancreatic alpha cells, the α–TC1 subclone 6 (α–TC1-6) glucagonoma cell line was chosen as a suitable model as it secretes glucagon exclusively, compared with other α–cell lines (such as α–TC3) which also secrete low levels of insulin (Hohmeier & Newgard, 2004). Qualitative determination of PANDER expression in these cells using immunofluorescent labeling revealed a diffuse, often peripheral profile with some similarity to the granular, mostly peripheral pattern of glucagon staining observed here typical of α–cell lines (Fig. 1A – C). While some overlap of PANDER and glucagon is observed by wide-field microscopy (Fig. 1C, inset), specific sub-cellular localization of PANDER within the α–TC1-6 cell could not be determined using this approach. Nonetheless, immunofluorescent labeling demonstrates that PANDER is robustly expressed in this α–cell line. As expected, similarly labeled PANDER-null C2C12 murine myoblast cells stained negatively for both PANDER and glucagon (Fig. 1D – F).
To assess PANDER expression in intact α–cells in-situ, we performed sequential immunohistochemical / immunofluorescent labeling of paraffin-embedded murine pancreatic sections for PANDER and glucagon respectively. This approach was employed to overcome limitations of antibody sensitivity for endogenous PANDER detection by immunofluorescence in sectioned tissue. We observed robust PANDER labeling (yellow) within both the glucagon-negative, centrally located β–cells, and the peripheral population of glucagon-positive (magenta) α–cells indicated by overlap of PANDER and glucagon signals (Fig. 1G – I).
With qualitative evidence of PANDER expression in both α–cell lines and primary islets, quantitative assessment was performed, using a TaqMan® RT-PCR approach to measure pander mRNA in α–TC1-6, β–TC3, and the C2C12 cell lines. Similar levels of pander transcript were detected in the α–TC1-6 and β–TC3 cell lines, with no expression observed in C2C12 cells (Fig. 2A). Additionally, at the protein level, PANDER expression was evaluated by densitometric analysis of western immunoblots of lysates harvested from the α–TC1-6 and β–TC3 cells. When normalized to β–actin, we observed similar levels of PANDER in these islet cell lines (Fig. 2B). We note however, that there is a non-statistically significant trend to lower PANDER protein expression in the β–TC3 cell line as compared with α–TC1-6 cells, despite highly similar mRNA content.
Islets comprise roughly 1% of total pancreatic area, with β–cells representing 70 – 75%, α–cells 15 – 20%, and δ and ε–cells comprising the remaining <10% of islet mass. The generation of enriched populations of primary murine islet α–, β–, and non–α, non–β cells by different sorting techniques is therefore challenging, with maximum cellular yields of about 10% of starting material. The low number of α–cells obtained by islet sorting therefore precluded reliable assessment of PANDER protein expression in these primary cells. However, using the technique described by Pipeleers in 1985 (Pipeleers et al., 1985a), pander mRNA in islet cell fractions enriched in α–, β–, and non–α, non–β cells was measured. Using this approach we observe that pander is more highly expressed in the α–cell enriched population compared with both the β–cell and non–α, non–β cell enriched fractions. This approximate 6-fold difference is statistically significant (P < 0.05) (Fig. 2C). We also observed significantly higher PANDER expression in the non–α non–β cell enriched fraction compared to the β–cell population, most likely due to contamination of this fraction with residual α–cells from the sorting procedure. TaqMan® RT-PCR detection of insulin and glucagon transcripts in the three populations indicates the expected enrichment with glucagon and insulin expression predominantly confined to the α–cell, and β–cell enriched fractions respectively (P<0.0001 for both) (Fig. 2C). Overall, the data suggest that PANDER is similarly expressed in β– and α–cell lines, but more robustly expressed at the mRNA level in primary α–cells as compared with β–cells of murine islets.
The sub-cellular localization of secretory proteins, including hormones, provides insight into both regulatory mechanisms and biological function. As PANDER regulation is largely uncharacterized in α–cells, localization within α–TC1-6 cells specifically with respect to glucagon was investigated. Currently available PANDER antibodies lack the sensitivity required for discrete sub-cellular analysis. In order to circumvent this problem and facilitate precise PANDER detection, a C-terminal PANDER-3-FLAG fusion protein construct was generated (Fig. 3A).
Expression of the fusion protein in lysates of transiently transfected α–TC1-6 cells 24 hrs post-transfection was demonstrated by immunoblot (Fig. 3B). Both cleaved and un-cleaved forms of the fusion protein are detected in the lysates of transfected cells using both FLAG and PANDER polyclonal antibodies with absence of signal in the lysates of non-transfected cells (Fig. 3B). Secretion of the fusion protein into culture media of transfected cells was separately confirmed (data not shown).
To evaluate discrete PANDER localization, simultaneous co-labeling of PANDER-FLAG nucleofected α–TC1-6 cells for FLAG, glucagon, and various markers for sub-cellular compartments was performed (Fig. 3C – N). Based on nucleofection efficiency; only about 40% of cultured cells expressed the construct at sufficient levels to observe robust FLAG labeling. It is not surprising therefore that within a typical cell cluster not all cells label positively for FLAG expression. The markers used include Chromogranin-A, for neuroendocrine vesicle identification, GM-130 to label the cis-Golgi network, and protein disulfide isomerase (PDI), which labels the lumen of the endoplasmic reticulum (ER). Confocal micrographs demonstrate that while both PANDER-FLAG and glucagon localize to granular cytoplasmic compartments, minimal overlap of the two is observed (Fig. 3C – E). Modest overlap between FLAG and chromogranin-A is observed in these cells (Fig. 3F – H), however co-localization is observed between PANDER-FLAG and GM130 (cis-Golgi) (Fig. 3I – K), and PDI (endoplasmic reticulum) (Fig. 3L – N). Together, these data suggest that PANDER localizes predominantly to a glucagon-negative granular cytoplasmic compartment downstream of its processing in the Golgi and ER.
To further characterize PANDER in α–cells, the regulation of PANDER secretion in response to glucose, amino acids, and insulin was investigated. These nutrient and hormonal cues have been demonstrated to represent important α–cell secretagogues and inhibitors respectively (Gromada et al., 2007). Electrogenic entry of positively charged amino acids (such as arginine) through cationic transporters induces membrane depolarization, thus triggering glucagon release (Gromada et al., 2007). Due to the technical limitations of detecting and reliably measuring secreted PANDER, it was necessary to examine PANDER release using time-points determined by previous studies at which significant accumulation in culture media is expected (Yang et al., 2005). When stimulated with a dose-curve of arginine over 16 hrs, we observe robust stimulation of PANDER secretion from transfected α–TC1-6 cells at 5 mM, becoming and remaining statistically significant between 10 and 20 mM arginine (Fig. 4A). While there appears to be an influence of arginine on PANDER content, the slight variability observed in cell lysate PANDER expression more directly reflects the transfection efficiency, as expression is not driven by the endogenous PANDER promoter. The doublet observed on immunoblot for PANDER reflects alternate cleavage sites within the signal peptide of full-length PANDER. As such, both bands of the doublet are considered in densitometric analysis. As a control for cell function, secreted glucagon was also measured (Fig. 4B). We observed significant glucagon release in response to 15 and 20 mM arginine, with non-significant stimulation at lower concentrations. In attempting to characterize arginine-stimulated PANDER secretion over time, we observed consistent basal secretion beginning at 2 through 24 hrs of continuous stimulation and identified a significant effect of arginine on PANDER release by two-way ANOVA (Arginine, P<0.004; Time, P<0.0002), (Fig. 4C). It is difficult to comment on the significance of this observation as we lack suitable methods of detecting secreted PANDER at early time-points. Earlier detection of PANDER, perhaps within 15 – 20 minutes post-stimulation would be ideal, but is not currently feasible using currently available antibodies. This may however reflect rapid release of PANDER in response to arginine challenge, consistent with the proposed mechanism of arginine induced exocytosis as via membrane depolarization. Despite minimal co-localization of PANDER and glucagon, both are positively regulated by arginine, indirectly indicating localization of PANDER to a secretory compartment.
One of the most physiologically relevant stimuli influencing α–cells is glucose, with hypoglycemia inducing significant glucagon release (Gromada et al., 2007). Whether this effect is entirely direct, or partially indirect via de-repression as β–cell paracrine effector levels decline in-vivo remains elusive. However, α–TC1-6 cells have been shown to secrete glucagon in response to acute hypoglycemia (Hohmeier & Newgard, 2004). When exposed to increasing or decreasing concentrations of glucose, basal PANDER secretion from transfected α–TC1-6 cells remains unchanged, without any stimulatory or inhibitory effect on PANDER observed (data not shown).
Interesting roles are currently debated for β–cell secretory products in the regulation of α–cell secretion. Chief among these factors is the glucagon antagonist insulin. Insulin is proposed to have a direct inhibitory effect on glucagon secretion from α–cells, particularly in the context of elevated local glucose concentration (Bailey et al., 2007;Xu et al., 2006). In addressing a potential inhibitory effect of insulin, on arginine-stimulated PANDER release, we exposed PANDER-transfected α–TC1-6 cells to 17 nM insulin, 20 mM L-arginine or a combination of both. Insulin significantly stimulated PANDER release, with apparent synergy observed in combination with arginine (Fig. 5A). In these studies there was a concordant increase in both PANDER and glucagon secretion and in response to 20 mM L-arginine. Interestingly, insulin had a robust stimulatory effect on PANDER release, and we observed an expected inhibition of glucagon secretion. In combination, insulin effectively suppressed arginine-induced glucagon release, while stimulated PANDER secretion remained.
With the unexpected positive regulation of PANDER by low levels of insulin, we focused our investigations on elucidating the mechanism by which insulin achieves this effect. Insulin exerts its metabolic roles by signaling through the insulin receptor (IR) complex via phosphoinositide-3-kinase (PI3K) and the Akt/PKB node (Taniguchi et al., 2006). We therefore assessed the requirement for signaling through PI3K and Akt for insulin-regulated PANDER secretion. We employed both biochemical and genetic approaches. Initially, we assessed PANDER secretion in the presence of the specific PI3K inhibitor wortmannin. When pre-treated with wortmannin, insulin-stimulated Akt phosphorylation at Ser 473, was significantly attenuated compared with control lysates, as expected (Fig. 5B). Insulin-regulated PANDER release was significantly reduced in stimulated cells pre-treated with wortmannin as compared with control cells (Fig. 5C). These results demonstrate that signaling through PI3K is essential for insulin-regulated PANDER release.
To further assess the requirement for signaling through the insulin receptor complex via the Akt node for insulin-stimulated PANDER secretion, we employed α–TC6 cells stably expressing a lentivirally-encoded anti-sense hairpin targeted to the β–chain of the insulin receptor. These cells were generated and characterized by the Wheeler group at the University of British Colombia (Diao et al., 2005). These cells have reduced insulin receptor expression (Diao et al., 2005) which we confirmed (data not shown), and in response to insulin-challenge demonstrate reduced Akt activation as compared with control cells (Fig. 6A) (Diao et al., 2005). On exposure to a similar dose-curve of insulin as used for wortmannin-treated cells, insulin-induced PANDER secretion was significantly attenuated in α–TC-IR cells, particularly at 17 nM (Fig. 6B). At 17 nM the insulin-stimulated increase no longer achieved statistical significance, while at 50 nM, secretion is diminished as compared with control α–TC1-6 cells (Fig. 6B). Therefore, using both genetic and biochemical approaches, we have demonstrated that signaling through IR–β via PI3K and Akt is necessary for insulin-regulated PANDER secretion from α–cells.
Overall, our report characterizing the expression, distribution, and regulation of PANDER in pancreatic α–cells indicates robust expression and localization of this novel protein to a granular compartment discrete from glucagon, which is positively regulated in response to insulin stimulation.
Despite considerable progress in the characterization of PANDER, the biological function remains unknown. Current evidence indicates a putative role in regulating glucose homeostasis. Indeed, the transcriptional activity and secretion of PANDER protein from the pancreatic β–cell are dose-dependently regulated by glucose (Burkhardt et al., 2005;Yang et al., 2005;Wang et al., 2008). The predominant restriction of PANDER to the glucagon- and insulin-positive cells of the pancreatic islet, the chief organ of gluco-regulation, provides indication of its potential role (Zhu et al., 2002). Furthermore, both PANDER mouse models (PANDER transgenic – TG, and knockout – PANDER −/−) reveal multiple gluco-regulatory aberrations (Burkhardt et al., 2009;Robert et al., 2006).
The islet-specific PANDER over-expressing transgenic mouse model (TG) displays high-fat-induced glucose intolerance, elevated hepatic glucose production, and impaired hepatic insulin signaling. Targeted disruption of PANDER in a global knockout mouse model results in impairment of glucose tolerance, stimulated insulin release, and diminished hepatic glucose production (Burkhardt et al., 2009). Recent evidence also demonstrates that PANDER binds to liver membrane and attenuates proximal steps of the insulin signaling cascade in human liver-derived HepG2 cells (Yang et al., 2009). Taken together, the various studies suggest that PANDER may serve as a novel islet-derived hormone which acts distally at the liver to modulate insulin action. However, to further understand how PANDER exerts its peripheral effects, it is necessary to provide additional biological characterization with regard to its expression and regulated secretion. Our present study provides further clues addressing regulated PANDER release from the previously neglected α–cell population.
Secretion from pancreatic α–cells is robustly stimulated by decreasing glucose concentrations (hypoglycemia) in-vivo, amino acid challenge, islet paracrine effectors, and other neuroendrocrine stimuli (Gromada et al., 2007). In contrast, elevated glucose and insulin concentrations are proposed to individually and synergistically suppress glucagon release, with potential roles attributed to novel β–cell factors such as γ–aminobutyric acid (GABA). Our studies demonstrate distinct localization of PANDER and glucagon within α–cells, which is concordant with the observed differential regulation of the two. Increasing and decreasing glucose concentrations exerts no influence on PANDER release, while there is a robust stimulatory effect of L-arginine. Further contrasting with glucagon regulation is insulin-induced PANDER secretion from α–cells, which is dependent on signaling through the insulin receptor complex downstream of PI3K and Akt/PKB. The distinct localization of PANDER and glucagon, coupled with insulin-induced PANDER release and targeting to the liver, is consistent with a role for PANDER in the modulation of hepatic insulin activity and glucose homeostasis (Yang et al., 2009).
Greater pander mRNA expression within α–cells as compared to β–cells suggests that the biological role of PANDER may be more relevant to these cells. All neuroendocrine cells including pancreatic α– and β–cells contain at least two types of secretory vesicles: the dense-core glucagon and insulin-containing granules respectively, and smaller synaptic-like microvesicles (SLMV’s) (Moriyama et al., 1996;Thomas-Reetz & De, 1994;Regnier-Vigouroux & Huttner, 1993). In β–cells, SLMV’s contain GABA, while in α–cells they appear to contain glutamate as suggested by the presence of membrane-bound glutamate transporters (Gaskins et al., 1995;Llona, 1995;Wiedenmann et al., 1988). The regulated secretion of both GABA and glutamate is emerging as a potential modulatory mechanism for islet paracrine function with peripheral effects on glucose homeostasis (Xu et al., 2006;Cabrera et al., 2008;Bertrand et al., 1992;Bertrand et al., 1993;Moriyama et al., 2000;Hayashi et al., 2003;Llona, 1995). While the regulation of SLMV’s has been more thoroughly investigated in β–cells than in α-cells, the exact nature and contents of α–cell SLMV’s are largely unknown. It has been shown in β–cells that despite significant responses to the same secretagogues there is differential regulation of GABA-containing SLMV’s compared to the dense-core insulin granules. The mechanisms by which these microvesicles are regulated are not well understood, but appear to involve both constitutive and regulated secretory pathways. The regulatory similarities and critical differences observed with PANDER and glucagon are concordant with their discrete localization within the pancreatic α–cell, and both basal and stimulated PANDER release. Taken together, this implies the potential localization of PANDER to SLMV’s within these cells.
The current lack of sensitive in-house and commercially available PANDER antibodies has hindered our ability to adequately detect endogenous PANDER protein. This has warranted the use of both transient transfection for over-expression and epitope (FLAG)-tagging of PANDER for stringent detection by western analysis and confocal immunofluorescent microscopy, respectively. This approach for characterizing both protein expression and localization has been extensively utilized by a number of groups without demonstrating protein mis-localization due to over-expression or presence of the (FLAG) epitope (Bhattacharyya et al., 1995;Olkkonen et al., 1993;Lipman et al., 1998;Duan et al., 1995;Advani et al., 1998). By employing this approach we are therefore confident that the expression of the fusion protein and its discrete localization within a granular compartment separate from glucagon in α–TC1-6 cell has been successfully demonstrated.
Pancreatic islets are highly vascularized and innervated, allowing their constituent cells to respond cooperatively to achieve and maintain normoglycemia in the balancing of energy demand and nutrient availability. While the factors influencing glucagon are not well understood, a number of nutritional, hormonal, and neural stimuli have been implicated (Gerich et al., 1976;Pipeleers et al., 1985b). There is also considerable evidence supporting the paracrine/endocrine theory of α–cell function, underscoring the influence of β–cell secretory products on glucagon release (Gromada et al., 2007). Given the disparate localization of glucagon and PANDER, it is not surprising that the two are differentially regulated by insulin. While signaling through PI3K and Akt has been shown to be inhibitory to glucagon release (Diao et al., 2005), this signal cascade robustly induces PANDER secretion. Once released into the circulation, PANDER likely targets distal sites such as the liver to exert its normal biological function (Yang et al., 2009). As a subtle modulator of hepatic insulin action, PANDER released from α–cells may act in lieu of or cooperatively with glucagon to enhance hepatic glucose production.
Progression to overt type 2 diabetes occurs via increasing peripheral insulin resistance, and overt disease is marked by hyperinsulinemia resulting from β–cell compensation, and hyperglucagonemia due to α–cell dysfunction, concomitant with the hyperglycemia resulting from decreased insulin sensitivity, increased glucagon action and endogenous glucose production. The relevance of α–cell function and hepatic glucose production to chronic hyperglycemia is progressively emerging as a critical component in our understanding of the underlying pathologies. What is highly intriguing yet requires further investigation is that various parameters of metabolic syndrome such as hyperglycemia, and hyper-insulinemia have now been shown to induce PANDER secretion from pancreatic β– and α–cells, respectively. Coupled with a potential role in modulating insulin action, it is reasonable to speculate that PANDER may possibly serve as a co-factor in the onset and subsequent progression of metabolic syndrome by exacerbating hepatic insulin resistance.
In summary, we have demonstrated the robust expression of PANDER within pancreatic α–cells and its localization to a discrete vesicular cytoplasmic compartment apart from glucagon. We have also elucidated both nutritional and hormonal regulation of PANDER release from α–cells. We observe no effect of glucose on PANDER secretion, while both L-arginine and insulin are robustly stimulatory. Further, insulin-stimulated PANDER secretion is dependent on signaling through the insulin receptor, downstream of PI3K and via the Akt node. These and other data from previous studies together suggest that PANDER released into systemic circulation may exert distal effects at the liver to modulate insulin action and hepatic glucose production, thereby influencing glycemia.
The authors would to acknowledge Dr. Edward Williamson and the Pathology Core of the Children’s Hospital of Philadelphia for invaluable assistance with confocal imaging studies. This publication was made possible through core services provided by the DERC at the University of Pennsylvania (DK 19525). This work was supported by grant K01-DK070744 (to B.R.B.) from the NIDDK, National Institutes of Health, and the Juvenile Diabetes Research Foundation (to B.A.W.).
This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases (DK070744 to B.R.B.) and the Juvenile Diabetes Research Foundation (JDRF).
The material contained in this manuscript was used for the partial fulfillment of the Ph.D. dissertation requirement for J.R.C. at the University of Pennsylvania in the laboratory of Dr. Bryan Wolf.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As aservice to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.