- Email: [email protected]
Establishment of monolayer culture of pig pancreatic endocrine cells by use of nicotinamide
Diabetes Research and Clinical Practice 42 (1998) 1 – 8
Establishment of monolayer culture of pig pancreatic endocrine cells by use of nicotinamide Hisako Ohgawara a,*, Tomohide Shikano b, Kazuyosi Fukunaga a, Mayumi Yamagishi a, Shunichi Miyazaki a,b,c a
Medical Research Institute, Tokyo Women’s Medical Uni6ersity, School of Medicine, 8 -1 Kawada-cho, Shinjuku-ku, Tokyo 162, Japan b Department of Physiology, Tokyo Women’s Medical Uni6ersity, School of Medicine, 8 -1 Kawada-cho, Shinjuku-ku, Tokyo 162 -8666, Japan c Laboratory of Intracellular Metabolism, Department of Molecular Physiology, National Institute for Physiological Science, 38 Aza-saigounaka, Myoudai-cho, Okazaki-shi, Shizuoka 44, Japan Received 6 May 1998; received in revised form 25 June 1998; accepted 19 August 1998
Abstract A method for the isolation and primary monolayer culture of adult pig pancreatic endocrine (PE) cells was established. Cells were dissociated from the pancreas by autodigestion without addition of proteolytic enzymes and separated into distinct bands in a single centrifugation step using Histopaque-1077 (a mixture of polysucrose and sodium diatrizoate). The cells collected from an interfacial fraction were suspended in RPMI 1640 containing 11 mmol/l D-glucose with or without nicotinamide (0, 10, 20, 40 mmol/l), and then placed in culture dishes. Pancreatic cells formed a monolayer while fibroblasts became detached from the bottom of the dish when cultured in the presence of nicotinamide. More than 80% of monolayer-forming cells were stained for insulin, using an enzymatic method, and were identified as B-cells. Morphologically, the PE cells extended multiple processes terminating in growth-cone-like structures, as visualized by both light microscopy and scanning electron microscopy. Insulin secretion in response to glucose stimulation occurred for 35 days of incubation in the RPMI 1640 medium, with or without nicotinamide. Exposure of the cells to nicotinamide for 35 days resulted in a 2 – 3-fold increase in insulin secretion in response to high glucose stimulus (16.7 mmol/l) compared with low glucose (5.5 mmol/l). Glucose-induced Ca2 + responses were examined in individual cells cultured for 35 days in the presence of 10 mmol/l nicotinamide, using Ca2 + imaging with fura-2. These results indicate that it is possible to prepare pig PE cells in monolayer culture with low fibroblast contamination and to maintain functioning B-cells in vitro for relatively long periods. The present method provides useful preparations for further morphological and physiological studies on the differentiation, growth and regenerative capacity of islet cells. © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. Keywords: Monolayer culture; Nicotinamide; Pig pancreatic endocrine cells * Corresponding author. Tel.: +81 3 33538111; fax: +81 3 52697364; e-mail: [email protected] 0168-8227/98/$ - see front matter © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. PII: S0168-8227(98)00096-5
2
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
1. Introduction In the near future, clinical programs of islet transplantation will depend upon sources of large mammals, such as pigs [1 – 4]. Porcine islets have immunological properties similar to human pancreatic islets, and the structure of porcine insulin differs only by a single amino acid from that of human insulin. Therefore, pig pancreatic endocrine (PE) cells possess a potential clinical applicability as a bioartificial pancreas when appropriately encapsulated [5], or placed into a hollow fiber [6] and diffusion chamber [7]. Pig PE cells are useful analogues for studies of the biochemical and physiological properties of human B-cells. We have previously reported a simple method for the preparation of adult pig PE cells by autodigestion of the pancreas without using proteolytic enzymes [8]. Attempts have also been made to purify adult pig PE cells [9]. As the next step, we report here a method for the formation of monolayer cultures of PE cells by the use of nicotinamide. Our results indicate that nicotinamide maintains a monolayer-forming cell culture of PE cells from adult pig while preventing the overgrowth of fibroblasts. The PE cells cultured for relatively long periods were investigated in terms of morphological changes and glucose-induced Ca2 + responses and insulin secretion.
sels in the tissues were discarded. Phosphatebuffered saline (PBS) containing 3.0% newborn calf serum (NCS) and 5.5 mmol/l D-glucose was injected into the pancreatic tissues, which were then minced with scissors. The pieces were transferred to a steel filter-basket with a mesh of 0.7 mm and a stirring bar (diameter 3.0 cm). The basket was then placed in a beaker which contained 250 ml of the cold PBS containing 3.0% NCS and 5.5 mmol/l D-glucose. The minced tissue was stirred gently (130 rpm/min) for 1 min, and the solution was discarded. Cold PBS was added, the mixture was gently stirred for 10 min, and the solution was collected from the beaker. This procedure was repeated twice. The mixture in the beaker was then gently stirred for 15 min with a magnetic bar (7.0 cm) instead of the stirring bar. The procedure was repeated at least four times, during which the small pieces of pancreas were
2. Materials and methods
2.1. Preparation of the pancreas The outline of the method for preparation and purification of adult pig PE cells without the addition of proteolytic enzyme is illustrated in Fig. 1. Adult pig pancreases were obtained from a local slaughterhouse. After the pancreas was exposed to 2.0% providone – isodine solution for 30 – 50 s, it was transported to the laboratory in cold (4°C) RPMI 1640 containing 10 mmol/l nicotinamide. The pancreas was immediately trimmed by careful dissection of surrounding fat tissue, lymph nodes, vessels and membranes on a large ice-cold plate. Serous membranes and ves-
Fig. 1. Schematic representation of the methods of preparation and purification of adult pig PE cells.
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
digested, presumably by endogenous proteases, and pancreatic cell clusters were easily separated from the connective tissue fragments by gentle stirring. The digested pancreatic cell clusters were passed through the steel filter, and the cells and cell clusters were collected immediately by centrifugation for 1 min at 3000 rpm. The resulting pellet of cells was resuspended in cold PBS solution and centrifuged at 1200 rpm for 2 min. This procedure was repeated four times, and each time the supernatant was collected. The harvest of pellets (approximately 15.0 ml in each pellet) was resuspended in PBS, and was placed onto 10.0 ml of Histopaque-1077 (a mixture of polysucrose and sodium diatrizoate; density 1.00790.001; Sigma Co., Ltd., Los Angeles, USA), and centrifuged at 1800 rpm for 10 min at room temperature, giving rise to one distinct band. The cells at each interface were collected and suspended in cold RPMI 1640 containing 10% heat-inactivated fetal calf serum. An aliquot of cells was stained with dithizone [10] and the cells were counted.
2.2. Cell culture Each aliquot (4.0×104) of cells was placed into a 35-mm culture dish (Sumilon cell desk LF, 30.0-mm diameter plastic coverslip in a culture dish; Sumitomo Bakelite Co., Ltd., Tokyo, Japan). The culture dishes were divided into four groups for a comparative study of the dose effect of nicotinamide on cell survival. Glucose-stimulated insulin secretion was measured on days 7 and 35 of culture in RPMI 1640 containing 10% fetal bovine serum (FBS) with or without nicotinamide (10, 20, 40 mmol/l). During the culture period, the cells were refed with 2.0 ml of the medium every 3 or 4 days. On days 7 and 35 of culture, the medium was discarded and replaced with 1.0 ml of the RPMI 1640 medium containing 3.3 mmol/l D-glucose and 10% FBS. The cells were incubated for 60 min at 37°C in 5% CO2 and air. Then the medium was replaced with either 1.0 ml of RPMI 1640 containing 5.5 mmol/l D-glucose (low glucose) or 16.7 mmol/l D-glucose (high glucose) and 10% FBS. The secretory responsiveness of the cells to glucose (5.5 or 16.7
3
mmol/l D-glucose) was determined during a 90min incubation at 37°C in 5% CO2 and air. After the incubation, the medium was collected and stored at − 20°C for insulin assay using an antibody bead insulin radioimmunoassay kit (Eiken Chemical Co., Ltd., Tokyo, Japan). The secretory experiment (static incubation) was repeated five times. On day 35, the cells were examined for their structural and functional characteristics. Structural integrity was assessed by cell viability, and by enzymatic staining and electron microscopy.
2.3. Microscopic obser6ations On days 10 and 35 of culture, the cells were rinsed with PBS and fixed with 4.0% paraformaldehyde in 0.1 mol/l phosphate buffer (pH 7.4) for 6 h. Immunoperoxidase staining for insulin was performed. The fixed cells were rinsed with PBS and incubated in normal goat serum (1:10) for 30 min, followed by treatment with anti-insulin guinea-pig serum (1:1000; DAKO, Denmark) for 4 h. They were then treated with horseradish peroxidase-labelled anti-guinea-pig serum (1:100; DAKO, Denmark) for 30 min. The reaction products were visualized with 0.01% 3,3%diaminobenzidine and 0.03% H2O2. The cells were photographed through a light microscope. For scanning electron microscopy at day 35 of culture with 10 mmol/l nicotinanide, the cells were fixed with 2.0% glutaraldehyde in PBS for 4 h. After rinsing with PBS, the cells were post-fixed with 1.0% osmium tetroxide for 1 h, and dehydrated with ethanol. The cells were then dried, coated with platinum–palladium in an ion coater (E-1030, Hitachi) and observed under a scanning electron microscope (S-800, Hitachi) with an accelerating voltage of 3 or 5 kV.
2.4. [Ca 2 + ]i measurement: To examine glucose-induced Ca2 + responses, purified PE cells were cultured in glass-bottomed plastic dishes (Glass Bottom; Iwaki Glass Co., Tokyo, Japan). A 8-mm hole was made at the bottom of the dish and a glass coverslip (12 mm diameter, 0.17 mm thickness) was attached from
4
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
Fig. 2. Photomicrographs of PE cells purified using Histopaque-1077 and stained with dithizone. Most of the cells are positively stained with dithizone. Fig. 3. Light-microscopic photographs of monolayer-forming adult pig pancreatic cells that were cultured in different culture media with or without nicotinamide. In A (for 10 days) and B (for 35 days) the cells were cultured in medium RPMI 1640 containing 10% FBS without nicotinamide. In C, the cells were cultured in medium RPMI 1640 supplemented with 10 mmol/l nicotinamide and 10% FBS for 35 days. In D and E, the cells were cultured in a medium containing nicotinamide 20 – 40 mmol/l and 10% FBS. Fig. 6. Ca2 + responses in the monolayer-forming pancreatic cells on day 35 of culture, before and after the addition of high glucose (16.7 mmol/l D-glucose). The color scale indicates Ca2 + concentration: blue corresponds to low calcium levels and red corresponds to high levels.
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
underneath to allow passage of UV light. The glass was treated with 5 ng/ml poly-L-lysine and 40 mg/ml mouse laminin aqueous solution for 35 h to facilitate adherence of the cells to the glass. On day 35 of culture, cells were loaded with the Ca2 + -sensitive dye fura-2 by incubation in RPMI medium containing 2 mM fura-2-AM (acetoxymethyl derivative; Molecular Probes, Eugene, OR) and 0.04% detergent pluronic F-127 (Molecular Probes) for 30 min. Ca2 + responses to D-glucose (3.3 or 16.7 mmol/l) were monitored in single cells by measuring the fluorescence emitted by fura-2. For activation of fura-2 fluorescence, UV light of 340 nm or 380 nm wavelength (UV340 or UV380) was produced by a xenon lamp and a 3409 10 or 380910 nm narrow band-pass filter and applied to the cell through a ×40 objective lens. Fura-2-loaded cells were irradiated every 1 or 2 min by UV340 for 0.25 s, followed 1 s later by a 0.25-s exposure to UV380. Emission fluorescence (F) was led to a silicon intensifier target camera through the same objective lens and a band-pass filter of 510 910 nm. Images were accumulated by an image processor (Arugus 50; Hamamatsu Photonics, Hamamatsu, Japan). After the experiments, the data were processed to obtain the ratio R =F340/F380. A calibration curve between R and Ca2 + concentration was obtained by measuring R of standard Ca2 + solutions containing N-(2-hydroxyethyl)ethylenedinitrilotriacetic acid.
3. Results Fig. 2 shows a photomicrograph of the cells collected from an interfacial fraction using Histopaque. Most of the cells were positively stained with dithizone. The percentage of dithizonestained cells was 94.6+2.5% (mean+ S.E.M., seven experiments), indicating that the fraction was rich in B-cells. Thus, the collected cells can be referred to as PE cells. Effects of addition of nicotinamide to the culture medium on PE cells and fibroblasts were investigated by immunohistochemical staining for
5
Fig. 4. Scanning electron micrograph of cultured monolayer cells on day 35 in a medium containing 10 mmol/l nicotinamide. Note the numerous microvilli on the cell surfaces showing a neuron-like differentiation ( ×3700). Some of cells terminated in cone-like structures (inset, × 5400).
insulin on days 10 and 35. In the absence of nicotinamide, the PE cells remained round and did not adhere to the bottom of the dish (brown cells in Fig. 3A; day 10). On day 35, the culture dish was occupied by fibroblasts, whereas the PE cells were hardly seen (Fig. 3B). Most of the PE cells might have been washed away during the medium exchange, while the fibroblasts overgrew during culture. In the presence of nicotinamide (10, 20, 40 mmol/l), the PE cells adhered to the bottom of the dish, while fibroblasts became detached. The PE cells began to extend multiple processes within 3 days of culture, and then formed a monolayer. On day 35, the monolayer consisted mainly of insulinproducing B-cells and fibroblasts were scarce in the dish (Fig. 3C–E) The higher the nicotinamide concentration, the lower the proprtion of fibroblasts (Fig. 3C,D). Thus, nicotinamide facilitates adherence of the B cells to the bottom of the culture dish and monolayer formation, while it prevents overgrowth of fibroblasts. The B-cells attained a star-like shape during culture in the presence of nicotinamide, with multiple processes. Scanning electron microscopy (Fig. 4) showed that the cell body was extended in several directions and that a few thin processes
6
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
had formed. A triangular-shaped small swelling was seen in the middle of each thin process. The end of the thin process was swollen, like the growth cone of neurons.
4. Discussion In the present study, we succeeded in purifying pig PE cells and a culture of B-cells functioning in vitro for a relatively long period. Previously, we
3.1. Insulin secretory responses To examine whether the cultured B-cells possess normal functions, the insulin secretory response to glucose was observed on days 7 and 35. When the cells were maintained in the medium without nicotinamide, no significant difference between stimulation by 5.5 and 16.7 mmol/l D-glucose on day 7 (Fig. 5A) and day 35 (Fig. 5B, left) could be observed. In contrast, when cells were cultured in the presence of 10 or 20 mmol/l nicotinamide, the insulin secretion from the cells exposed to high glucose (16.7 mmol/l) was two to three times higher than that from the cells stimulated by low glucose (5.5 mmol/l) (Fig. 5B, middle). In the cells maintained in the medium containing 40 mmol/l nicotinamide, insulin secretion was not enhanced by high glucose concentration. On day 70, only the cells that were cultured in the presence of 10 mmol/l nicotinamide still possessed the ability to secrete insulin when challenged with high glucose. Thus, moderate doses of nicotinamide, particularly 10 mmol/l, were favorable for maintaining glucose-induced insulin secretory responses of cultured PE cells (data not shown).
3.2. Ca 2 + responses to glucose stimulation It is well known that B-cells exhibit an increase in [Ca2 + ]i in response to D-glucose as the prerequisite for exocytosis of insulin [11]. Glucose-induced Ca2 + responses were examined in individual cells cultured for 35 days in the presence of 10 mmol/l nicotinamide, using Ca2 + imaging with fura-2. Fig. 6A shows the effect on [Ca2 + ]i of a change in glucose concentration from 3.3 to 16.7 mmol/l D-glucose, in the medium containing 10 mmol/l nicotinamide on day 35 of culture. The [Ca2 + ]i measured in single cells of pancreatic islets undergoes oscillations that are caused by glucose-induced burst of electrical activity. In the presence of high glucose, the [Ca2 + ]i reached 50–300 mM.
Fig. 5. Basal and glucose-stimulated insulin secretion from the pancreatic cells following culture in different media with or without nicotinamide. (A) Cells cultured for 7 days in RPMI 1640 containing 10% FBS. (B) Cells cultured in RPMI 1640, with or without nicotinamide (0, 10, 20 or 40 mmol/l) for 35 days. On the following day, insulin secretion was measured after a 90-min stimulation by a low or high dose of glucose, as described in Section 2. Open columns ( ) show the insulin secretion by stimulation with 5.5 mmol/l D-glucose, and shaded columns ( ) show the insulin release after stimulation with 16.7 mmol/l D-glucose. Data are the mean9S.E.M. of four replications from three separate experiments. * P B0.05, ** P B0.01 compared to control using Student’s two-sample t-test.
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
reported a simple method for purification of adult pig PE cells using the MPRM (mono-poly resolving medium) system [9]. The cell population included cells stained by antibodies against insulin (60.0 –67.7%), for glucagon (45.8 – 52.3%) and for somatostatin (6.9–9.8%). The proportion of Bcells purified by the present method with Histopaque was substantially higher (94.5%), and the harvested cell number was greater than that obtained by the previous method. It is well known that survival of monolayerforming cell cultures of PE cells is only minimal because of the overgrowth of fibroblasts. In this study, we demonstrated that in a primary culture of adult pig PE cells in a medium containing nicotinamide fibroblast overgrowth was reduced or completely eliminated. Recently, Hiromatu et al. [12] investigated the effects of nicotinamide on cultured orbital fibroblasts from patients with thyroid- associated ophthalmopathy. They reported that nicotinamide inhibited fibroblast proliferation as assessed by [3H]thymidine incorporation and cell-counting [13]. In our experiments, high doses of nicotinamide (30–40 mmol/l) strongly reduced adhesion of fibroblastic cells to the culture dish, compared to low doses (10 mmol/l), indicating the dose-dependent inhibition of fibroblast outgrowth. At the high dose of nicotinamide (30 – 40 mmol/l), monolayer-forming B-cells gradually became functionally exhausted. With the moderate dose of nicotinamide (10 mmol/l), cells survived well, maintaining their ability for glucose-stimulated insulin secretion. The mechanism of action of nicotinamide has been attributed to inhibition of the nuclear enzyme, poly(ADP-ribose)synthetase, which, if activated during DNA repair synthesis, could lead to a critical decrease in the NAD level of the B-cells [13]. Interestingly, nicotinamide induced morphological changes of cultured PE cells to a neuronlike structure. The insulin-secreting B-cells resemble neurons in a number of ways, although they are not derived from neuron ectoderm [14] but instead arise from endoderm [15]. For example, B-cells have been shown to contain small neurosecretory vesicles in addition to insulin-secreting granules [16]. The neurotransmitter acetyl-
7
choline depolarizes the plasma membrane of B-cells, allowing influx of Ca2 + , and initiates the exocytosis of insulin-containing granules [17]. Pancreatic B-cells also resemble neurons in that they express some neurotransmitter biosynthetic enzymes, such as tyrosine hydroxylase, DOPA decarboxylase and glutamate decarboxylase (an autoimmune target in type 1 diabetes mellitus), as well as neurofilament protein and the neural cell adhesion molecule N-CAM [18–21]. The characteristic changes in [Ca2 + ]i were observed in monolayer-forming PE cells in response to 16.7 mmol/l glucose. Oscillations in [Ca2 + ]i have been shown in single islets of the rat [22] and mouse [23,24] and in single B-cells [25,26] upon stimulation with glucose. Recently, Westerlund et al. [27] demonstrated a relationship between insulin release and [Ca2 + ]i in single islets from ob/ob mice by use of hypoglycemic sulfonylurea in the presence of low (3.0 mmol/l) or high (11 mmol/l) glucose. The results showed that pulsatile insulin release did not always depend on [Ca2 + ]i oscillations. It was suggested that cyclic generation of ATP might fuel pulsatile release under conditions where [Ca2 + ]i remains stable. In summary, we have described a procedure for preparing a high yield of normal adult pig B-cells that retain their ability to secrete insulin in response to glucose, and can be maintained in a monolayer without fibroblasts. This procedure provides a relatively inexpensive source of B-cells for transplantation studies as well as for biochemical, physiological and molecular biological studies of B-cell function.
Acknowledgements We are grateful to Dr D.F. Steiner, University of Chicago, and B. Levene for their comments on this manuscript. This study was supported, in part, by grants from the Japan Society for the Promotion of Science, Research for the Future (JSPS-RFTF 96100201), Creative Basic Research from the Ministry of Education, Science, Sports and Culture (10NP0201) and Research on Advanced Medical Technology.
8
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
References [1] C. Ricordi, C. Socci, A. Davalli, C. Staudacher, P. Baro, A. Vertova, I. Sassi, F. Gavazzi, G. Pozza, V. Dicarlo, Isolation of the eluvive pig islet, Surgery 107 (1990) 688 – 694. [2] R. Calafiore, F. Calcinaro, G. Basta, M. Pietropaolo, A. Falorni, M. Piermattei, P. Brunett, A method for massive separation of highly purified, adult porcine islets of Langerhans, Metabolism 39 (1990) 175–181. [3] B.J. Hering, D. Roman, A. Clarius, M. Brendel, M. Slijepcevic, G. Bretzel, K. Federlin, Xenogeneic islet preparation and transplantion. Bovine islet of Langerhans: potential souce for transplantation, Diabetes 38 (Suppl. 1) (1989) 206–208. [4] P.E. Lacy, E.T. Lacy, E.H. Finke, Y. Yasunami, An improved method for the isolation of islets from the beef pancreas, Diabetes 30 (Suppl. 4) (1982) 109–111. [5] F. Lim, A.M. Sun, Microcapsulated islet as bioartificial endocrine pancreas, Science 210 (1980) 908–910. [6] P.E. Lacy, O.D. Hegre, A.G. Vazeou, Maintenance of normoglycemia in diabetic mice by subcutaneous xenografts of encapsulated islets, Science 254 (1991) 1782. [7] H. Ohgawara, J.-I. Miyazaki, S. Karibe, N. Karagiri, F. Tasiro, T. Akaike, Assessment of pore size of a semipermeable membrane for immunoisolation on xenoimplantation of pancreatic B-cells using a diffusion chamber, Transplant. Proc. 27 (1995) 3319–3320. [8] H. Ohgawara, A. Kobayashi, S.-J. Chong, T. Akaike, Y. Hashimoto, Preparation of adult pig pancreatic cells: comparative study of methods with or without proteolytic enzymes, Cell Transplant. 3 (1994) 325–331. [9] S. Sato, H. Ohgawara, N. Katagiri, Y. Nakagawa, E. Aikawa, Y. Omori, K. Ota, A novel, simple method for purification of adult pig pancreatic endocrine cells, J. Tokyo Women’s Med. Coll. 66 (1996) 471–479. [10] Z.A. Latif, J. Nobe, R. Alejandro, A simple method of staining fresh and cultured islets, Transplantation 45 (1988) 827 – 830. [11] P. Bergsten, E. Grapengiesser, E. Gylfe, A. Tengholm, B. Hellman, Synchronous oscillations of cytoplasmic Ca2 + and insulin release in glucose-stimulated pancreatic islets, J. Biol. Chem. 269 (1994) 8749–8753. [12] Y. Hiromatsu, G. Gen, M. Migita, J. Kamachi, Immunomodulatory effect of nicotinamide on cultured human orbital fibroblasts from patients with thyroid-associated ophthalmopathy, Thyroid 5 (Suppl. 1) (1995) S17. [13] H. Okamoto, Molecular basis of experimental diabetes: degeneration, oncogenesis and regeneration of pancreatic B-cells of islets of Langerhans, Bioessays 2 (1984) 15–21. [14] A.G. Pearse, Islet cell precursors are neurons, Nature 295 (1982) 96 – 97.
.
[15] J. Fontaine, C. Le Lievre, N.M. Le Douarin, What is the developmental fate of the neural crest cells which migrate into the pancreas in the avian embryo?, Gen. Comp. Endocrinol. 33 (1977) 394 – 404. [16] P. De Camilli, R. Jahn, Pathways to regulated exocytosis in neurons, Annu. Rev. Physiol. 52 (1986) 625 – 645. [17] B. Wiedenmann, W.W. Franke, C. Kuhn, R. Moll, V.E. Gould, Synaptophysin: a marker protein for neuroendocrine cells and neoplasms, Proc. Natl. Acad. Sci. USA 83 (1986) 3500 – 3504. [18] G. Teitelman, J.K. Lee, Cell lineage analysis of pancreatic islet development: glucagon and insulin cells arise from catecholaminergic precursors present in the pancreatic duct, Dev. Biol. 121 (1987) 454 – 466. [19] G.S. Eisenbarth, K. Shimizu, M.A. Bowring, S. Wells, Expression of receptors for tetanus toxin and monoclonal antibody A2B5 by pancreatic islet cells, Proc. Natl. Acad. Sci. USA 79 (1982) 5066 – 5070. [20] S. Baekkeskov, H.J. Aanstoot, S. Christgau, A. Reetz, M. Solimena, M. Cascalho, F. Folli, H. Richter-Olesen, P. DeCamilli, Identification of the 64K autoantigen in insulin-dependent diabetes as the GADA-synthesizing enzyme glutamic acid decarboxylase, Nature 347 (1990) 151 – 156. [21] O.K. Langley, M.C. Aletsee-Ufrecht, N.J. Grant, M. Gratzl, Expression of the neural cell adhesion molecule NCAM in endocrine cells, J. Histol. Cytol. 37 (1989) 781 – 791. [22] L.M. Rosario, I. Atwater, A.M. Scott, Pulsatile insulin release and electrical activity from single ob/ob mouse islets of Langerhans, Adv. Exp. Med. Biol. 211 (1986) 413 – 425. [23] E. Longo, K. Tomheim, J. Deeney, B. Vamum, D. Tilloston, M. Prentki, B. Corkey, Oscillations in cytosolic free Ca2 + , oxygen consumption, and insulin secretion in glucose stimulated rat pancreatic islets, J. Biol. Chem. 266 (1991) 9314 – 9319. [24] M. Valdeolmillos, R.M. Santos, D. Contreras, B. Soria, L.M. Rosario, Glucose-induced oscillation Ca2 + concentration resembling bursting activity in single mouse islets of Langerhans, FEBS Lett. 259 (1989) 19 – 23. [25] M. Valdeolmillos, A. Nadal, B. Soria, J. Garcia-Sancho, Fluorescence digital image analysis of glucose-induced [Ca2 + ]i oscillation in mouse pancreatic islets of Langerhans, Diabetes 42 (1993) 1210 – 1214. [26] W.F. Pralong, C. Bartley, C.B. Wollheim, Single islet B-cell stimulation by nutrients: relationship between pyridine nucleotides, cytosolic Ca2 + and secretion, EMBO J. 9 (1990) 53 – 60. [27] J. Westerlund, E. Gylfe, P. Bergsten, Pulsatile insulin release from pancreatic islets with nonoscillatory elevation of cytoplasmic Ca2 + , J. Clin. Invest. 100 (1997) 2547 – 2551.
Establishment of monolayer culture of pig pancreatic endocrine cells by use of nicotinamide Hisako Ohgawara a,*, Tomohide Shikano b, Kazuyosi Fukunaga a, Mayumi Yamagishi a, Shunichi Miyazaki a,b,c a
Medical Research Institute, Tokyo Women’s Medical Uni6ersity, School of Medicine, 8 -1 Kawada-cho, Shinjuku-ku, Tokyo 162, Japan b Department of Physiology, Tokyo Women’s Medical Uni6ersity, School of Medicine, 8 -1 Kawada-cho, Shinjuku-ku, Tokyo 162 -8666, Japan c Laboratory of Intracellular Metabolism, Department of Molecular Physiology, National Institute for Physiological Science, 38 Aza-saigounaka, Myoudai-cho, Okazaki-shi, Shizuoka 44, Japan Received 6 May 1998; received in revised form 25 June 1998; accepted 19 August 1998
Abstract A method for the isolation and primary monolayer culture of adult pig pancreatic endocrine (PE) cells was established. Cells were dissociated from the pancreas by autodigestion without addition of proteolytic enzymes and separated into distinct bands in a single centrifugation step using Histopaque-1077 (a mixture of polysucrose and sodium diatrizoate). The cells collected from an interfacial fraction were suspended in RPMI 1640 containing 11 mmol/l D-glucose with or without nicotinamide (0, 10, 20, 40 mmol/l), and then placed in culture dishes. Pancreatic cells formed a monolayer while fibroblasts became detached from the bottom of the dish when cultured in the presence of nicotinamide. More than 80% of monolayer-forming cells were stained for insulin, using an enzymatic method, and were identified as B-cells. Morphologically, the PE cells extended multiple processes terminating in growth-cone-like structures, as visualized by both light microscopy and scanning electron microscopy. Insulin secretion in response to glucose stimulation occurred for 35 days of incubation in the RPMI 1640 medium, with or without nicotinamide. Exposure of the cells to nicotinamide for 35 days resulted in a 2 – 3-fold increase in insulin secretion in response to high glucose stimulus (16.7 mmol/l) compared with low glucose (5.5 mmol/l). Glucose-induced Ca2 + responses were examined in individual cells cultured for 35 days in the presence of 10 mmol/l nicotinamide, using Ca2 + imaging with fura-2. These results indicate that it is possible to prepare pig PE cells in monolayer culture with low fibroblast contamination and to maintain functioning B-cells in vitro for relatively long periods. The present method provides useful preparations for further morphological and physiological studies on the differentiation, growth and regenerative capacity of islet cells. © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. Keywords: Monolayer culture; Nicotinamide; Pig pancreatic endocrine cells * Corresponding author. Tel.: +81 3 33538111; fax: +81 3 52697364; e-mail: [email protected] 0168-8227/98/$ - see front matter © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. PII: S0168-8227(98)00096-5
2
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
1. Introduction In the near future, clinical programs of islet transplantation will depend upon sources of large mammals, such as pigs [1 – 4]. Porcine islets have immunological properties similar to human pancreatic islets, and the structure of porcine insulin differs only by a single amino acid from that of human insulin. Therefore, pig pancreatic endocrine (PE) cells possess a potential clinical applicability as a bioartificial pancreas when appropriately encapsulated [5], or placed into a hollow fiber [6] and diffusion chamber [7]. Pig PE cells are useful analogues for studies of the biochemical and physiological properties of human B-cells. We have previously reported a simple method for the preparation of adult pig PE cells by autodigestion of the pancreas without using proteolytic enzymes [8]. Attempts have also been made to purify adult pig PE cells [9]. As the next step, we report here a method for the formation of monolayer cultures of PE cells by the use of nicotinamide. Our results indicate that nicotinamide maintains a monolayer-forming cell culture of PE cells from adult pig while preventing the overgrowth of fibroblasts. The PE cells cultured for relatively long periods were investigated in terms of morphological changes and glucose-induced Ca2 + responses and insulin secretion.
sels in the tissues were discarded. Phosphatebuffered saline (PBS) containing 3.0% newborn calf serum (NCS) and 5.5 mmol/l D-glucose was injected into the pancreatic tissues, which were then minced with scissors. The pieces were transferred to a steel filter-basket with a mesh of 0.7 mm and a stirring bar (diameter 3.0 cm). The basket was then placed in a beaker which contained 250 ml of the cold PBS containing 3.0% NCS and 5.5 mmol/l D-glucose. The minced tissue was stirred gently (130 rpm/min) for 1 min, and the solution was discarded. Cold PBS was added, the mixture was gently stirred for 10 min, and the solution was collected from the beaker. This procedure was repeated twice. The mixture in the beaker was then gently stirred for 15 min with a magnetic bar (7.0 cm) instead of the stirring bar. The procedure was repeated at least four times, during which the small pieces of pancreas were
2. Materials and methods
2.1. Preparation of the pancreas The outline of the method for preparation and purification of adult pig PE cells without the addition of proteolytic enzyme is illustrated in Fig. 1. Adult pig pancreases were obtained from a local slaughterhouse. After the pancreas was exposed to 2.0% providone – isodine solution for 30 – 50 s, it was transported to the laboratory in cold (4°C) RPMI 1640 containing 10 mmol/l nicotinamide. The pancreas was immediately trimmed by careful dissection of surrounding fat tissue, lymph nodes, vessels and membranes on a large ice-cold plate. Serous membranes and ves-
Fig. 1. Schematic representation of the methods of preparation and purification of adult pig PE cells.
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
digested, presumably by endogenous proteases, and pancreatic cell clusters were easily separated from the connective tissue fragments by gentle stirring. The digested pancreatic cell clusters were passed through the steel filter, and the cells and cell clusters were collected immediately by centrifugation for 1 min at 3000 rpm. The resulting pellet of cells was resuspended in cold PBS solution and centrifuged at 1200 rpm for 2 min. This procedure was repeated four times, and each time the supernatant was collected. The harvest of pellets (approximately 15.0 ml in each pellet) was resuspended in PBS, and was placed onto 10.0 ml of Histopaque-1077 (a mixture of polysucrose and sodium diatrizoate; density 1.00790.001; Sigma Co., Ltd., Los Angeles, USA), and centrifuged at 1800 rpm for 10 min at room temperature, giving rise to one distinct band. The cells at each interface were collected and suspended in cold RPMI 1640 containing 10% heat-inactivated fetal calf serum. An aliquot of cells was stained with dithizone [10] and the cells were counted.
2.2. Cell culture Each aliquot (4.0×104) of cells was placed into a 35-mm culture dish (Sumilon cell desk LF, 30.0-mm diameter plastic coverslip in a culture dish; Sumitomo Bakelite Co., Ltd., Tokyo, Japan). The culture dishes were divided into four groups for a comparative study of the dose effect of nicotinamide on cell survival. Glucose-stimulated insulin secretion was measured on days 7 and 35 of culture in RPMI 1640 containing 10% fetal bovine serum (FBS) with or without nicotinamide (10, 20, 40 mmol/l). During the culture period, the cells were refed with 2.0 ml of the medium every 3 or 4 days. On days 7 and 35 of culture, the medium was discarded and replaced with 1.0 ml of the RPMI 1640 medium containing 3.3 mmol/l D-glucose and 10% FBS. The cells were incubated for 60 min at 37°C in 5% CO2 and air. Then the medium was replaced with either 1.0 ml of RPMI 1640 containing 5.5 mmol/l D-glucose (low glucose) or 16.7 mmol/l D-glucose (high glucose) and 10% FBS. The secretory responsiveness of the cells to glucose (5.5 or 16.7
3
mmol/l D-glucose) was determined during a 90min incubation at 37°C in 5% CO2 and air. After the incubation, the medium was collected and stored at − 20°C for insulin assay using an antibody bead insulin radioimmunoassay kit (Eiken Chemical Co., Ltd., Tokyo, Japan). The secretory experiment (static incubation) was repeated five times. On day 35, the cells were examined for their structural and functional characteristics. Structural integrity was assessed by cell viability, and by enzymatic staining and electron microscopy.
2.3. Microscopic obser6ations On days 10 and 35 of culture, the cells were rinsed with PBS and fixed with 4.0% paraformaldehyde in 0.1 mol/l phosphate buffer (pH 7.4) for 6 h. Immunoperoxidase staining for insulin was performed. The fixed cells were rinsed with PBS and incubated in normal goat serum (1:10) for 30 min, followed by treatment with anti-insulin guinea-pig serum (1:1000; DAKO, Denmark) for 4 h. They were then treated with horseradish peroxidase-labelled anti-guinea-pig serum (1:100; DAKO, Denmark) for 30 min. The reaction products were visualized with 0.01% 3,3%diaminobenzidine and 0.03% H2O2. The cells were photographed through a light microscope. For scanning electron microscopy at day 35 of culture with 10 mmol/l nicotinanide, the cells were fixed with 2.0% glutaraldehyde in PBS for 4 h. After rinsing with PBS, the cells were post-fixed with 1.0% osmium tetroxide for 1 h, and dehydrated with ethanol. The cells were then dried, coated with platinum–palladium in an ion coater (E-1030, Hitachi) and observed under a scanning electron microscope (S-800, Hitachi) with an accelerating voltage of 3 or 5 kV.
2.4. [Ca 2 + ]i measurement: To examine glucose-induced Ca2 + responses, purified PE cells were cultured in glass-bottomed plastic dishes (Glass Bottom; Iwaki Glass Co., Tokyo, Japan). A 8-mm hole was made at the bottom of the dish and a glass coverslip (12 mm diameter, 0.17 mm thickness) was attached from
4
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
Fig. 2. Photomicrographs of PE cells purified using Histopaque-1077 and stained with dithizone. Most of the cells are positively stained with dithizone. Fig. 3. Light-microscopic photographs of monolayer-forming adult pig pancreatic cells that were cultured in different culture media with or without nicotinamide. In A (for 10 days) and B (for 35 days) the cells were cultured in medium RPMI 1640 containing 10% FBS without nicotinamide. In C, the cells were cultured in medium RPMI 1640 supplemented with 10 mmol/l nicotinamide and 10% FBS for 35 days. In D and E, the cells were cultured in a medium containing nicotinamide 20 – 40 mmol/l and 10% FBS. Fig. 6. Ca2 + responses in the monolayer-forming pancreatic cells on day 35 of culture, before and after the addition of high glucose (16.7 mmol/l D-glucose). The color scale indicates Ca2 + concentration: blue corresponds to low calcium levels and red corresponds to high levels.
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
underneath to allow passage of UV light. The glass was treated with 5 ng/ml poly-L-lysine and 40 mg/ml mouse laminin aqueous solution for 35 h to facilitate adherence of the cells to the glass. On day 35 of culture, cells were loaded with the Ca2 + -sensitive dye fura-2 by incubation in RPMI medium containing 2 mM fura-2-AM (acetoxymethyl derivative; Molecular Probes, Eugene, OR) and 0.04% detergent pluronic F-127 (Molecular Probes) for 30 min. Ca2 + responses to D-glucose (3.3 or 16.7 mmol/l) were monitored in single cells by measuring the fluorescence emitted by fura-2. For activation of fura-2 fluorescence, UV light of 340 nm or 380 nm wavelength (UV340 or UV380) was produced by a xenon lamp and a 3409 10 or 380910 nm narrow band-pass filter and applied to the cell through a ×40 objective lens. Fura-2-loaded cells were irradiated every 1 or 2 min by UV340 for 0.25 s, followed 1 s later by a 0.25-s exposure to UV380. Emission fluorescence (F) was led to a silicon intensifier target camera through the same objective lens and a band-pass filter of 510 910 nm. Images were accumulated by an image processor (Arugus 50; Hamamatsu Photonics, Hamamatsu, Japan). After the experiments, the data were processed to obtain the ratio R =F340/F380. A calibration curve between R and Ca2 + concentration was obtained by measuring R of standard Ca2 + solutions containing N-(2-hydroxyethyl)ethylenedinitrilotriacetic acid.
3. Results Fig. 2 shows a photomicrograph of the cells collected from an interfacial fraction using Histopaque. Most of the cells were positively stained with dithizone. The percentage of dithizonestained cells was 94.6+2.5% (mean+ S.E.M., seven experiments), indicating that the fraction was rich in B-cells. Thus, the collected cells can be referred to as PE cells. Effects of addition of nicotinamide to the culture medium on PE cells and fibroblasts were investigated by immunohistochemical staining for
5
Fig. 4. Scanning electron micrograph of cultured monolayer cells on day 35 in a medium containing 10 mmol/l nicotinamide. Note the numerous microvilli on the cell surfaces showing a neuron-like differentiation ( ×3700). Some of cells terminated in cone-like structures (inset, × 5400).
insulin on days 10 and 35. In the absence of nicotinamide, the PE cells remained round and did not adhere to the bottom of the dish (brown cells in Fig. 3A; day 10). On day 35, the culture dish was occupied by fibroblasts, whereas the PE cells were hardly seen (Fig. 3B). Most of the PE cells might have been washed away during the medium exchange, while the fibroblasts overgrew during culture. In the presence of nicotinamide (10, 20, 40 mmol/l), the PE cells adhered to the bottom of the dish, while fibroblasts became detached. The PE cells began to extend multiple processes within 3 days of culture, and then formed a monolayer. On day 35, the monolayer consisted mainly of insulinproducing B-cells and fibroblasts were scarce in the dish (Fig. 3C–E) The higher the nicotinamide concentration, the lower the proprtion of fibroblasts (Fig. 3C,D). Thus, nicotinamide facilitates adherence of the B cells to the bottom of the culture dish and monolayer formation, while it prevents overgrowth of fibroblasts. The B-cells attained a star-like shape during culture in the presence of nicotinamide, with multiple processes. Scanning electron microscopy (Fig. 4) showed that the cell body was extended in several directions and that a few thin processes
6
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
had formed. A triangular-shaped small swelling was seen in the middle of each thin process. The end of the thin process was swollen, like the growth cone of neurons.
4. Discussion In the present study, we succeeded in purifying pig PE cells and a culture of B-cells functioning in vitro for a relatively long period. Previously, we
3.1. Insulin secretory responses To examine whether the cultured B-cells possess normal functions, the insulin secretory response to glucose was observed on days 7 and 35. When the cells were maintained in the medium without nicotinamide, no significant difference between stimulation by 5.5 and 16.7 mmol/l D-glucose on day 7 (Fig. 5A) and day 35 (Fig. 5B, left) could be observed. In contrast, when cells were cultured in the presence of 10 or 20 mmol/l nicotinamide, the insulin secretion from the cells exposed to high glucose (16.7 mmol/l) was two to three times higher than that from the cells stimulated by low glucose (5.5 mmol/l) (Fig. 5B, middle). In the cells maintained in the medium containing 40 mmol/l nicotinamide, insulin secretion was not enhanced by high glucose concentration. On day 70, only the cells that were cultured in the presence of 10 mmol/l nicotinamide still possessed the ability to secrete insulin when challenged with high glucose. Thus, moderate doses of nicotinamide, particularly 10 mmol/l, were favorable for maintaining glucose-induced insulin secretory responses of cultured PE cells (data not shown).
3.2. Ca 2 + responses to glucose stimulation It is well known that B-cells exhibit an increase in [Ca2 + ]i in response to D-glucose as the prerequisite for exocytosis of insulin [11]. Glucose-induced Ca2 + responses were examined in individual cells cultured for 35 days in the presence of 10 mmol/l nicotinamide, using Ca2 + imaging with fura-2. Fig. 6A shows the effect on [Ca2 + ]i of a change in glucose concentration from 3.3 to 16.7 mmol/l D-glucose, in the medium containing 10 mmol/l nicotinamide on day 35 of culture. The [Ca2 + ]i measured in single cells of pancreatic islets undergoes oscillations that are caused by glucose-induced burst of electrical activity. In the presence of high glucose, the [Ca2 + ]i reached 50–300 mM.
Fig. 5. Basal and glucose-stimulated insulin secretion from the pancreatic cells following culture in different media with or without nicotinamide. (A) Cells cultured for 7 days in RPMI 1640 containing 10% FBS. (B) Cells cultured in RPMI 1640, with or without nicotinamide (0, 10, 20 or 40 mmol/l) for 35 days. On the following day, insulin secretion was measured after a 90-min stimulation by a low or high dose of glucose, as described in Section 2. Open columns ( ) show the insulin secretion by stimulation with 5.5 mmol/l D-glucose, and shaded columns ( ) show the insulin release after stimulation with 16.7 mmol/l D-glucose. Data are the mean9S.E.M. of four replications from three separate experiments. * P B0.05, ** P B0.01 compared to control using Student’s two-sample t-test.
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
reported a simple method for purification of adult pig PE cells using the MPRM (mono-poly resolving medium) system [9]. The cell population included cells stained by antibodies against insulin (60.0 –67.7%), for glucagon (45.8 – 52.3%) and for somatostatin (6.9–9.8%). The proportion of Bcells purified by the present method with Histopaque was substantially higher (94.5%), and the harvested cell number was greater than that obtained by the previous method. It is well known that survival of monolayerforming cell cultures of PE cells is only minimal because of the overgrowth of fibroblasts. In this study, we demonstrated that in a primary culture of adult pig PE cells in a medium containing nicotinamide fibroblast overgrowth was reduced or completely eliminated. Recently, Hiromatu et al. [12] investigated the effects of nicotinamide on cultured orbital fibroblasts from patients with thyroid- associated ophthalmopathy. They reported that nicotinamide inhibited fibroblast proliferation as assessed by [3H]thymidine incorporation and cell-counting [13]. In our experiments, high doses of nicotinamide (30–40 mmol/l) strongly reduced adhesion of fibroblastic cells to the culture dish, compared to low doses (10 mmol/l), indicating the dose-dependent inhibition of fibroblast outgrowth. At the high dose of nicotinamide (30 – 40 mmol/l), monolayer-forming B-cells gradually became functionally exhausted. With the moderate dose of nicotinamide (10 mmol/l), cells survived well, maintaining their ability for glucose-stimulated insulin secretion. The mechanism of action of nicotinamide has been attributed to inhibition of the nuclear enzyme, poly(ADP-ribose)synthetase, which, if activated during DNA repair synthesis, could lead to a critical decrease in the NAD level of the B-cells [13]. Interestingly, nicotinamide induced morphological changes of cultured PE cells to a neuronlike structure. The insulin-secreting B-cells resemble neurons in a number of ways, although they are not derived from neuron ectoderm [14] but instead arise from endoderm [15]. For example, B-cells have been shown to contain small neurosecretory vesicles in addition to insulin-secreting granules [16]. The neurotransmitter acetyl-
7
choline depolarizes the plasma membrane of B-cells, allowing influx of Ca2 + , and initiates the exocytosis of insulin-containing granules [17]. Pancreatic B-cells also resemble neurons in that they express some neurotransmitter biosynthetic enzymes, such as tyrosine hydroxylase, DOPA decarboxylase and glutamate decarboxylase (an autoimmune target in type 1 diabetes mellitus), as well as neurofilament protein and the neural cell adhesion molecule N-CAM [18–21]. The characteristic changes in [Ca2 + ]i were observed in monolayer-forming PE cells in response to 16.7 mmol/l glucose. Oscillations in [Ca2 + ]i have been shown in single islets of the rat [22] and mouse [23,24] and in single B-cells [25,26] upon stimulation with glucose. Recently, Westerlund et al. [27] demonstrated a relationship between insulin release and [Ca2 + ]i in single islets from ob/ob mice by use of hypoglycemic sulfonylurea in the presence of low (3.0 mmol/l) or high (11 mmol/l) glucose. The results showed that pulsatile insulin release did not always depend on [Ca2 + ]i oscillations. It was suggested that cyclic generation of ATP might fuel pulsatile release under conditions where [Ca2 + ]i remains stable. In summary, we have described a procedure for preparing a high yield of normal adult pig B-cells that retain their ability to secrete insulin in response to glucose, and can be maintained in a monolayer without fibroblasts. This procedure provides a relatively inexpensive source of B-cells for transplantation studies as well as for biochemical, physiological and molecular biological studies of B-cell function.
Acknowledgements We are grateful to Dr D.F. Steiner, University of Chicago, and B. Levene for their comments on this manuscript. This study was supported, in part, by grants from the Japan Society for the Promotion of Science, Research for the Future (JSPS-RFTF 96100201), Creative Basic Research from the Ministry of Education, Science, Sports and Culture (10NP0201) and Research on Advanced Medical Technology.
8
H. Ohgawara et al. / Diabetes Research and Clinical Practice 42 (1998) 1–8
References [1] C. Ricordi, C. Socci, A. Davalli, C. Staudacher, P. Baro, A. Vertova, I. Sassi, F. Gavazzi, G. Pozza, V. Dicarlo, Isolation of the eluvive pig islet, Surgery 107 (1990) 688 – 694. [2] R. Calafiore, F. Calcinaro, G. Basta, M. Pietropaolo, A. Falorni, M. Piermattei, P. Brunett, A method for massive separation of highly purified, adult porcine islets of Langerhans, Metabolism 39 (1990) 175–181. [3] B.J. Hering, D. Roman, A. Clarius, M. Brendel, M. Slijepcevic, G. Bretzel, K. Federlin, Xenogeneic islet preparation and transplantion. Bovine islet of Langerhans: potential souce for transplantation, Diabetes 38 (Suppl. 1) (1989) 206–208. [4] P.E. Lacy, E.T. Lacy, E.H. Finke, Y. Yasunami, An improved method for the isolation of islets from the beef pancreas, Diabetes 30 (Suppl. 4) (1982) 109–111. [5] F. Lim, A.M. Sun, Microcapsulated islet as bioartificial endocrine pancreas, Science 210 (1980) 908–910. [6] P.E. Lacy, O.D. Hegre, A.G. Vazeou, Maintenance of normoglycemia in diabetic mice by subcutaneous xenografts of encapsulated islets, Science 254 (1991) 1782. [7] H. Ohgawara, J.-I. Miyazaki, S. Karibe, N. Karagiri, F. Tasiro, T. Akaike, Assessment of pore size of a semipermeable membrane for immunoisolation on xenoimplantation of pancreatic B-cells using a diffusion chamber, Transplant. Proc. 27 (1995) 3319–3320. [8] H. Ohgawara, A. Kobayashi, S.-J. Chong, T. Akaike, Y. Hashimoto, Preparation of adult pig pancreatic cells: comparative study of methods with or without proteolytic enzymes, Cell Transplant. 3 (1994) 325–331. [9] S. Sato, H. Ohgawara, N. Katagiri, Y. Nakagawa, E. Aikawa, Y. Omori, K. Ota, A novel, simple method for purification of adult pig pancreatic endocrine cells, J. Tokyo Women’s Med. Coll. 66 (1996) 471–479. [10] Z.A. Latif, J. Nobe, R. Alejandro, A simple method of staining fresh and cultured islets, Transplantation 45 (1988) 827 – 830. [11] P. Bergsten, E. Grapengiesser, E. Gylfe, A. Tengholm, B. Hellman, Synchronous oscillations of cytoplasmic Ca2 + and insulin release in glucose-stimulated pancreatic islets, J. Biol. Chem. 269 (1994) 8749–8753. [12] Y. Hiromatsu, G. Gen, M. Migita, J. Kamachi, Immunomodulatory effect of nicotinamide on cultured human orbital fibroblasts from patients with thyroid-associated ophthalmopathy, Thyroid 5 (Suppl. 1) (1995) S17. [13] H. Okamoto, Molecular basis of experimental diabetes: degeneration, oncogenesis and regeneration of pancreatic B-cells of islets of Langerhans, Bioessays 2 (1984) 15–21. [14] A.G. Pearse, Islet cell precursors are neurons, Nature 295 (1982) 96 – 97.
.
[15] J. Fontaine, C. Le Lievre, N.M. Le Douarin, What is the developmental fate of the neural crest cells which migrate into the pancreas in the avian embryo?, Gen. Comp. Endocrinol. 33 (1977) 394 – 404. [16] P. De Camilli, R. Jahn, Pathways to regulated exocytosis in neurons, Annu. Rev. Physiol. 52 (1986) 625 – 645. [17] B. Wiedenmann, W.W. Franke, C. Kuhn, R. Moll, V.E. Gould, Synaptophysin: a marker protein for neuroendocrine cells and neoplasms, Proc. Natl. Acad. Sci. USA 83 (1986) 3500 – 3504. [18] G. Teitelman, J.K. Lee, Cell lineage analysis of pancreatic islet development: glucagon and insulin cells arise from catecholaminergic precursors present in the pancreatic duct, Dev. Biol. 121 (1987) 454 – 466. [19] G.S. Eisenbarth, K. Shimizu, M.A. Bowring, S. Wells, Expression of receptors for tetanus toxin and monoclonal antibody A2B5 by pancreatic islet cells, Proc. Natl. Acad. Sci. USA 79 (1982) 5066 – 5070. [20] S. Baekkeskov, H.J. Aanstoot, S. Christgau, A. Reetz, M. Solimena, M. Cascalho, F. Folli, H. Richter-Olesen, P. DeCamilli, Identification of the 64K autoantigen in insulin-dependent diabetes as the GADA-synthesizing enzyme glutamic acid decarboxylase, Nature 347 (1990) 151 – 156. [21] O.K. Langley, M.C. Aletsee-Ufrecht, N.J. Grant, M. Gratzl, Expression of the neural cell adhesion molecule NCAM in endocrine cells, J. Histol. Cytol. 37 (1989) 781 – 791. [22] L.M. Rosario, I. Atwater, A.M. Scott, Pulsatile insulin release and electrical activity from single ob/ob mouse islets of Langerhans, Adv. Exp. Med. Biol. 211 (1986) 413 – 425. [23] E. Longo, K. Tomheim, J. Deeney, B. Vamum, D. Tilloston, M. Prentki, B. Corkey, Oscillations in cytosolic free Ca2 + , oxygen consumption, and insulin secretion in glucose stimulated rat pancreatic islets, J. Biol. Chem. 266 (1991) 9314 – 9319. [24] M. Valdeolmillos, R.M. Santos, D. Contreras, B. Soria, L.M. Rosario, Glucose-induced oscillation Ca2 + concentration resembling bursting activity in single mouse islets of Langerhans, FEBS Lett. 259 (1989) 19 – 23. [25] M. Valdeolmillos, A. Nadal, B. Soria, J. Garcia-Sancho, Fluorescence digital image analysis of glucose-induced [Ca2 + ]i oscillation in mouse pancreatic islets of Langerhans, Diabetes 42 (1993) 1210 – 1214. [26] W.F. Pralong, C. Bartley, C.B. Wollheim, Single islet B-cell stimulation by nutrients: relationship between pyridine nucleotides, cytosolic Ca2 + and secretion, EMBO J. 9 (1990) 53 – 60. [27] J. Westerlund, E. Gylfe, P. Bergsten, Pulsatile insulin release from pancreatic islets with nonoscillatory elevation of cytoplasmic Ca2 + , J. Clin. Invest. 100 (1997) 2547 – 2551.