Prolactin Protects Against Diabetes Induced by Multiple Low Doses of Streptozotocin in Mice
229 Prolactin protects against diabetes induced by multiple low doses of streptozotocin in mice
M Holstad and S Sandler Department of Medical Cell Biology, Uppsala University, PO Box 571, SE–751 23 Uppsala, Sweden (Requests for offprints should be addressed to M Holstad, Department of Medical Cell Biology, Biomedicum, PO Box 571, SE–751 23 Uppsala, Sweden; Email: [email protected])
Abstract In earlier studies it has been shown that prolactin (PRL) is in vitro were studied. Mice treated with STZ became a stimulating factor for the immune system, and it has been gradually hyperglycemic, and concomitant treatment with suggested that PRL might antagonize immunosuppressive PRL (4 mg/kg body weight) for 21 days significantly effects of glucocorticoids. PRL has been reported to affect reduced the elevation in blood glucose levels from day 10 the cytokine secretion pattern, by elevating cytokine gene onwards (P<0·05). Morphologic examinations of the pan- expression in macrophages, after the onset of sepsis. It also creas on day 21 of mice receiving STZ injections revealed promotes the antibody response in mice where it increases a marked insulitis, but only moderate insulitis in the STZ the production of interferon- (IFN-) and inhibits treated animals given PRL. BC administration (10 mg/kg interleukin-1 (IL-1) production. Due to these properties, body weight) in combination with STZ did not signifi- PRL might influence the development of autoimmune cantly affect the elevation in blood glucose levels or the type 1 diabetes. The aim of the present study was to insulitis. PRL or BC administration alone did not change examine the effects of two drugs; PRL and bromocriptine the serum glucose concentration. This study indicates that (BC) in vivo on the development of hyperglycemia and PRL may affect hyperglycemia in the early phase of pancreatic insulitis in mice treated with multiple doses of autoimmune diabetes. We suggest that it might be due to streptozotocin (STZ) (40 mg/kg body weight, i.p.). The counteraction of autoimmune immunologic mechanisms dopaminergic agonist BC is known to inhibit PRL se- and/or enhancement of -cell regeneration. cretion. In another set of experiments, the direct effects of Journal of Endocrinology (1999) 163, 229–234 PRL on the function of pancreatic islets exposed to STZ
Introduction progression of the inflammatory response. Moreover, several experimental studies and clinical observations have Prolactin (PRL), a neuroendocrine peptide hormone suggested that PRL might be important in the pathogen- (24 kDa single structure) produced by the pituitary gland, esis of various autoimmune diseases (Ferrari et al. 1983, has been suggested to play an important role in the Berczi 1993, Walker et al. 1993, Neidhart 1998, regulation of the immune system because it stimulates Velkeniers et al. 1998), but very few studies have focused lymphocyte proliferation and macrophage function on the role of PRL in diabetes mellitus type 1. It is (Spangelo et al. 1987, Gala 1991). For example, PRL therefore interesting to study whether PRL affects has been reported to cause a significant elevation of diabetes induced by multiple low doses of streptozotocin cytokine gene expression in macrophages, after the (STZ), a well established animal model for the patho- onset of sepsis (Zhu et al. 1997). The stimulatory effect of genesis of type 1 diabetes (Like & Rossini 1976, Kolb- PRL on immunity may result from antagonism of the Bachofen et al. 1988). immunosuppressive glucocorticoid effects (Berczi 1997). The aim of the experiments in the present study was to PRL promotes the antibody response and increases examine the effects of two drugs; PRL and bromocriptine the production of interferon- (IFN-) and inhibits (BC) in vivo on the development of hyperglycemia and interleukin-1 (IL-1) production (Berczi 1997). A severe pancreatic insulitis in mice treated with multiple intra- inflammation often results in elevated levels of circulating peritoneal injections of STZ. The dopaminergic agonist glucocorticoids and adrenocorticotropin (ACTH), which BC is known to inhibit PRL secretion (Sinha et al. 1976). in turn, reduces the secretion of PRL (Bateman et al. 1989, In another set of experiments, the direct effects of PRL Parrott & Goode 1993). Administration of PRL, under on the function of pancreatic islets exposed to STZ these circumstances, might therefore influence the in vitro was studied. Previous data have shown that PRL
Journal of Endocrinology (1999) 163, 229–234 Online version via http://www.endocrinology.org 0022–0795/99/0163–229 1999 Society for Endocrinology Printed in Great Britain 230 M HOLSTAD and S SANDLER · Effects of PRL on type 1 diabetes
stimulates insulin secretion and proliferation of -cells in The pancreatic sections were evaluated by two inde- islets of murine and human origin (Brelje et al. 1993, pendent examiners being unaware of the origin of the Sorenson et al. 1993). sections.
Islet isolation, culture and incubation Materials and Methods Pancreatic islets were isolated from overnight fasted adult Animals and treatment C57BL/KsJ mice by collagenase digestion (Collagenase A, Boehringer-Mannheim, Mannheim, Germany), as Inbred adult male C57BL/KsJ mice, aged 3–4 months, described in detail elsewhere (Sandler et al. 1987). The weighing about 30 g, were used. These mice were orig- islets where then picked by hand, using a braking pipette. inally obtained from the Jackson Laboratory (Bar Harbor, Groups of 150–200 islets were precultured free-floating for ME, USA) but they have been locally bred at the 6–7 days at 37 C in an atmosphere of humidified air +5% Biomedical Centre, Uppsala, Sweden for more than 18 CO2 in 5 ml medium RPMI 1640 containing 11·1mM years. The animals had free access to tap water and glucose (Sigma) supplemented with 10% (v/v) fetal calf pelleted food throughout the experiments. ff serum (FCS) (Sigma). Medium was changed every second The mice were allocated to one of eight di erent groups day. After the preculture period, islets in groups of 50 were of treatment: saline+saline, saline+vehicle, saline+PRL, transferred to new culture dishes containing 1 ml Krebs– saline+BC, STZ+saline, STZ+vehicle, STZ+PRL, Ringer bicarbonate buffer+Hepes (KRBH; 114·3mM STZ+BC. PRL (luteotropic hormone from sheep NaCl, 4·74 mM KCl, 1·15 mM KH2PO4,1·18 mM pituitary glands) and STZ were purchased from Sigma MgSO ,25·0 mM NaHCO ,10·0 mM Hepes, 4·26 mM Chemicals (St Louis, MO, USA) and BC was kindly 4 3 NaOH, 2·54 mM CaCl2;pH7·4) supplemented with provided by Novartis (Ann Lindquist, Taby, Sweden). 2 mg/ml bovine serum albumin (BSA; fraction V; ICN PRL was dissolved in saline, and for BC a vehicle Biochemicals, Aurora, OH, USA) and 5·6 mM glucose composed of ethanol (70%):saline (1:9) and tartaric acid with or without addition of PRL (6 or 60 µg/ml). The (10 mg/kg body weight) was used. The mice received first islets were incubated at 37 C in a gas atmosphere of 95% either an intraperitoneal injection of saline (0·2 ml) or of O +5% CO for 30 min. Then, STZ, dissolved in 0·9% STZ (40 mg/kg body weight; 0·2 ml). STZ was dissolved 2 2 ff NaCl, was added to the incubation medium (final con- in cold citrate bu er (10 mM, pH 4·5). After 30 min the centration 1·8 mM STZ) for 30 min. After rinsing and mice were given a second intraperitoneal injection of removal of the medium the islets were cultured overnight either saline, vehicle or of PRL (4 mg/kg body weight; in RPMI 1640 supplemented with 10% FCS. 0·2 ml) or BC (10 mg/kg body weight; 0·2 ml). The first type of injection (saline or STZ) was given daily for 5 consecutive days, whilst the second type of injection Islet glucose-stimulated insulin release, insulin and DNA (saline, vehicle, PRL or BC) was given daily until the content animals were killed. After the culture period, triplicate groups of 10 islets were Blood glucose determinations (ExachTech blood glu- transferred to sealed glass vials containing 250 µl KRBH cose meter, Baxter Travenol, Deerfield, IL, USA) were supplemented with 2 mg/ml BSA and 1·7 mM glucose performed on day 0, before any injection and on days 3, 7, during the first hour. The islets were incubated at 37 Cin 10, 14, 17 and 21. The blood samples were obtained from a gas phase of 95% O2+5% CO2. The incubation medium a tail vein of non-fasted mice. was then gently removed and replaced by KRBH and 2 mg/ml BSA supplemented with 16·7 mM glucose dur- ing the second hour. After the incubations, the islets were Morphologic examination harvested and pooled in groups of 30 and homogenized in The mice were killed by cervical dislocation and their 0·2 ml redistilled water. A fraction of the homogenate was pancreatic glands were removed and fixed in 10% formalin mixed with acid ethanol and insulin was extracted over- solution and embedded in paraffin. Sections 7 µm thick night at 4 C. The insulin concentration of the extract was were cut and stained with hematoxylin and eosin. The measured by RIA (Heding 1972). Another fraction of the pancreatic islet histology was ranked according to four aqueous homogenate was used for measurement of DNA arbitrary classes as previously described and illustrated content (Hinegardner 1971). ( Jansson & Sandler 1988, Holstad & Sandler 1993); class A: normal islet structure, class B: mononuclear cell infil- Statistical analysis tration in the periphery of the islets, class C: infiltration of mononuclear cells into a majority of the islets i.e. insulitis, The blood glucose values are expressed as means.. and class D: only a few islets present often with altered groups of data were compared using Student’s unpaired structure, e.g. pyknotic cell nuclei. t-test. In the in vitro experiments, where islets were
Journal of Endocrinology (1999) 163, 229–234 Effects of PRL on type 1 diabetes · M HOLSTAD and S SANDLER 231
Figure 2 Blood glucose concentrations in C57BL/KsJ mice treated Figure 1 Blood glucose concentrations in C57BL/KsJ mice treated with i.p. injections of either STZ+vehicle (open squares; n=6) or with i.p. injections of either saline+saline ( ; n=5) or saline+PRL STZ+BC (diamonds; n=6). STZ (40 mg/kg body weight; 0·2ml) (; n=5) or STZ+saline ( ; n=11) or STZ+PRL ( ; n=11). STZ was given for 5 consecutive days, whilst BC (10 mg/kg body (40 mg/kg body weight; 0·2 ml) was given for 5 consecutive days, weight; 0·2 ml) or vehicle (ethanol (70%):saline (1:9) and tartaric whilst PRL (4 mg/kg body weight; 0·2 ml) or saline (0·2ml) acid (10 mg/kg body weight); 0·2 ml) treatment was maintained treatment was maintained for 21 days. Day 0 denotes the value for 21 days. Values are meansS.E. for n animals. before any drug has been administered. Immediately thereafter the animals were injected as described above. Values are meansS.E. for n animals. *P<0·05 vs the STZ+PRL treated group, using Student’s unpaired t-test. morphology, compared with the saline-treated group of animals (Table 1). incubated in triplicate, a mean was calculated for each Effects of PRL on STZ induced islet dysfunction in vitro experimental group and considered as one separate obser- In this series of experiments, pancreatic islets were pre- vation. Furthermore, every islet observation represents exposed for 30 min at 5·6 mM glucose to different different mouse islet donors and values were expressed as means.. and groups of data were compared using Table 1 Effects of PRL and BC on islet morphology. Pancreatic islet Student’s paired t-test. histology rank in male C57BL/KsJ mice treated with one daily i.p. injection of STZ (40 mg/kg body weight; 0·2 ml) or saline 9 g/l; 0.2 ml) for 5 days. In addition, the mice were also given a second Results i.p. injection of either PRL (4 mg/kg body weight; 0·2 ml), saline, BC (10 mg/kg body weight; 0·2 ml) or vehicle (ethanol (70%): ff saline 1:9) and tartaric acid (10 mg/kg body weight); 0·2ml) E ects of PRL and BC on multiple low dose STZ-induced 30 min after the first type of injections and maintained daily for 21 diabetes days. The histology was ranked according to four arbitrary classes: A, normal islet structure; B, mononuclear cell infiltration in the The mice treated with multiple low doses of STZ became periphery of the islets; C, infiltration of mononuclear cells into a gradually hyperglycemic (Fig. 1). It was found that treat- majority of islets i.e. insulitis; D, only a few islets left, often with ment with PRL during the observation period reduced the altered histology, for example pyknotic cell nuclei level of hyperglycemia from day 10 onwards (P<0·05) (Fig. 1). BC administration in combination with STZ did Pancreatic islet histology rank ff Treatment ABCD not significantly a ect the elevation in blood glucose levels Day 1–5 Day 1–21 (Number of animals) (Fig. 2). Morphologic examinations of the pancreas on day 21 of mice receiving STZ injections plus saline revealed a Saline + Saline 5000 marked insulitis and structural changes of the islets (Table Saline + PRL5000 1). This appeared to be modulated to some extent in the STZ + Saline 1244 PRL treated animals (Table 1). The degree of insulitis STZ + PRL3431 ff (new set of experiments) seemed una ected by BC (Table 1). PRL (Fig. 1) or BC STZ + vehicle 0141 (data not shown) administration alone did not change the STZ + BC0024 serum glucose concentration or affect the pancreatic islet
Journal of Endocrinology (1999) 163, 229–234 232 M HOLSTAD and S SANDLER · Effects of PRL on type 1 diabetes
Figure 3 Glucose-stimulated insulin release at 1·7 mM (black bars) Figure 5 Insulin (hatched bars) and DNA content (black bars) of pancreatic islets exposed to STZ (1·8 mM) and/or PRL (6 or or 16·7 mM (hatched bars) glucose from mouse pancreatic islets exposed to STZ (1·8 mM) in the presence or absence of PRL at 60 mg/l). Values are means S.E.; n=6. *P<0·05 vs control islets two different concentrations (6 or 60 mg/l), as indicated. Bars are not exposed to STZ or PRL, using Student’s paired t-test. meansS.E. for six experiments. *P<0·05 vs the control group not exposed to STZ or PRL, using Student’s paired t-test. but no change in the islet DNA content. There was a trend that the islet insulin content after STZ exposure was increased after addition of PRL, but this effect did not attain statistical significance (Fig. 5).
Discussion
The aim of the present study was to investigate if PRL can affect the development of type 1 diabetes in mice treated with multiple low doses of STZ. Due to previously reported immunologic actions of PRL (Berczi 1993, Walker et al. 1993, Neidhart 1998) we hypothesized that PRL might potentiate the development of autoimmune type 1 diabetes in this animal model. However, our results indicate that PRL has a beneficial effect on hyperglycemia Figure 4 Relative glucose stimulation of insulin secretion and insulitis in the early phase of autoimmune diabetes. (16·7 mM/1·7 mM). Data are calculated from values in Fig. 3. The reason for this effect is unclear, but it could be that *P<0·05 vs control islets not exposed to STZ or PRL, using PRL in itself enhances -cell function during a period Student’s paired t-test. when the -cell mass is being reduced. Other possibilities are that PRL decreases the -cell toxic effects of STZ or that PRL counteracts the autoimmune mechanisms concentrations of PRL and then incubated for 30 min in leading to -cell destruction. 1·8 mM STZ. After incubation, the islets were cultured Our results show that in vitro exposure of mouse overnight in RPMI 1640+10% FCS, whereupon glucose- pancreatic islets to PRL (6 and 60 mg/l), could not stimulated insulin release of the islets was examined. The prevent the inhibitory effects of STZ on glucose- control islets increased their insulin secretion about 7-fold stimulated insulin secretion. This argues against the view at 16·7 mM glucose, as compared with the secretion at that the protective effect of PRL against multiple low dose 1·7 mM glucose (Figs 3 and 4). The islets incubated in the STZ induced diabetes would be mediated by a direct presence of the two different concentrations of PRL alone effect of PRL on STZ action. It is therefore unlikely that responded similarly to glucose compared with the controls. the effect of PRL on hyperglycemia and insulitis observed STZ caused a marked inhibition of glucose-stimulated in vivo may be attributed to direct molecular interactions insulin release, and this was not counteracted by PRL. The between PRL and STZ, or an interference with cellular islet DNA and insulin contents were not changed by the events leading to -cell damage inflicted by STZ. brief incubation with PRL (Fig. 5) Exposure to 1·8mM Previous data have shown that long-term exposure to STZ caused about 30% reduction of the insulin content, PRL stimulates insulin secretion and proliferation of
Journal of Endocrinology (1999) 163, 229–234 Effects of PRL on type 1 diabetes · M HOLSTAD and S SANDLER 233
-cells in islets of murine and human origin (Brelje et al. BC+saline or vehicle compared with the correspondingly 1993, Stout et al. 1997). The possible beneficial action of saline+saline or vehicle treated group of animals. Our PRL in the present study could thus be attributed to study did not show any clear effect of BC, which should PRL-induced stimulation of -cell function. However, antagonize the action of PRL, on the incidence of diabetes we currently did not observe that PRL stimulated insulin in mice treated with multiple low doses of STZ. The release in vitro after short-term incubation. Moreover, an in vivo dose and administration route (i.p.) of BC was accurate quantification of -cell replication in vivo in an selected based on previous reports (Sinha et al. 1976, animal with insulitis would be difficult to perform, due to Durant et al. 1995). Of course, we can not exclude that the presence of infiltrating immune cells within the these doses were inadequate to affect the immune system islets. Thus, we cannot exclude that PRL could have in the C57BL/KsJ mice, or that another administration exerted a positive influence on the -cell mass during the route would give a different result. Another interesting development of low dose STZ diabetes. finding was that we observed hair loss in some animals It has been reported that hypophysectomy in rodents given BC. This is a well known side effect of this drug in causes deficiencies in both humoral and cell-mediated humans and at least it shows that BC in this experiment immunity (Gala 1991, Reber 1993) and administration of was pharmacologically active. PRL to such animals restores their immune function. PRL In summary, results from our study and from others receptors have been identified on the cell membranes of suggest that PRL may have a beneficial effect in auto- white blood cells, and lymphocytes have been shown to immune type 1 diabetes. It is likely that this reflects a secrete a PRL-like substance (Gala 1991, Reber 1993, combined effect of PRL on both the neuroendocrine, the Neidhart 1998). Either too high or too low levels of PRL immune system and perhaps also on the regulation of may affect these receptors and due to feed-back regulation, glucose metabolism. immunosuppressive actions by PRL may result. It is likely that PRL acts as a signal, regulating the immune cells in the present experiments. Macrophages are present in the Acknowledgements pancreatic islets very early after multiple STZ injections in mice and this is obligatory for the subsequent development We are grateful for the excellent technical assistance by of hyperglycemia (Kolb-Bachofen et al. 1988). PRL may Margareta Engkvist, Astrid Nordin and Eva Tornelius. inhibit IL-1 production from the macrophages (Berczi This study was supported by grants from the Swedish 1997) and thereby promote an altered immune response Medical Research Council (72X–8273, 12P–10739), the by shifting the balance from a cell-mediated to a humoral Novo Nordic Fund, the Swedish Diabetes Association, the immune response, which has been proposed earlier to Family Ernfors Fund, the Swedish American Diabetes affect the outcome of the autoimmune process in type 1 Research Program and the Swedish Society for Medical diabetes (Rabinovitch 1993, 1994, Adorini et al. 1996). Research. In the literature, only a few experimental studies have focused on the possible relationship between PRL and the development of type 1 diabetes. First, in humans, some References protective effects of BC, a dopaminergic agonist decreas- ing the secretion of PRL (Sinha et al. 1976), on -cell Adorini L, Guery JC & Trembleau S 1996 Manipulation of the function was observed in diabetic patients treated for Th1/Th2 cell balance: an approach to treat human autoimmune several months (Atkison et al. 1990). Secondly, with regard diseases? Autoimmunity 23 53–68. Atkison PR, Mahon JL, Dupre J, Stiller CR, Jenner MR, Paul TL & to the experimental animal models of type 1 diabetes, Momah CI 1990 Interaction of bromocriptine and cyclosporin in divergent results exist. A protective effect of BC against insulin dependent diabetes mellitus: results from the Canadian open hyperglycemia was observed in diabetes prone female study. Journal of Autoimmunity 3 793–799. non-obese diabetic (NOD) mice (Hawkins et al. 1994) Bateman A, Singh A, KralT&SolomonS1989 The immune– hypothalamic–pituitary–adrenal axis. Endocrine Reviews 10 92–112. but, in contrast, the drug appeared to accelerate the onset Berczi I 1993 The role of prolactin in the pathogenesis of autoimmune of diabetes in NOD males and significantly increased the disease. Endocrine Pathology 4 178–195. rate of insulitis in both sexes (Durant et al. 1995). Berczi I 1997 Pituitary hormones and immune function. Acta Paediatrica 423 70–75. BC, as well as other D2-dopaminergic agonists, may also have various metabolic effects, for example they produce a Brelje TC, Scharp DW, Lacy PE, Ogren L, Talamantes F, Robertson M, Friesen HG & Sorenson RL 1993 Effect of homologous marked hyperglycemia in rodents (Durant et al. 1995). It placental lactogens, prolactins, and growth hormones on islet B-cell also inhibited insulin release from isolated mouse pancre- division and insulin secretion in rat, mouse, and human islets: atic islets in vitro, suggesting that insulin release may be implication for placental lactogen regulation of islet function during controlled by D -dopaminergic receptors on the -cells pregnancy. Endocrinology 132 879–887. 2 Durant S, Alves V, Coulaud J, El Hasnaoui A, Dardenne M & (El-Denshary et al. 1982). We could not confirm this Homo-Delarche F 1995 Attempts to pharmacologically modulate finding since no significant difference in serum glucose prolactin levels and type 1 autoimmune diabetes in the non-obese concentration was observed in mice receiving PRL or diabetic (NOD) mouse. Journal of Autoimmunity 8 875–885.
Journal of Endocrinology (1999) 163, 229–234 234 M HOLSTAD and S SANDLER · Effects of PRL on type 1 diabetes
El-Denshary EE, Ismail NA, Gagerman E, SehlinJ&Taljedal IB Rabinovitch A 1994 Immunoregulatory and cytokine imbalances in 1982 Bromocriptine and insulin secretion. Bioscience Reports 2 the pathogenesis of IDDM. Therapeutic intervention by immuno- 115–116. stimulation? Diabetes 43 613–621. Ferrari C, BoghenM&Paracchi A 1983 Thyroid autoimmunity in Reber PM 1993 Prolactin and immunomodulation. American Journal of hyperprolactinemic disorders. Acta Endocrinologica 104 35–41. Medicine 95 637–644. Gala RR 1991 Prolactin and growth hormone in the regulation of the Sandler S, AnderssonA&Hellerstrom C 1987 Inhibitory effects of immune system. Proceedings of the Society for Experimental Biology and interleukin 1 on insulin secretion, insulin biosynthesis and oxidative Medicine 198 513–527. metabolism of rat pancreatic islets. Endocrinology 121 1424–1431. Hawkins TA, Gala RR & Dunbar JC 1994 Prolactin modulates the Sinha YN, Salocks CB & VanderLaan WP 1976 A comparison of the incidence of diabetes in male and female NOD mice. Autoimmunity effects of CB–154 and lergotrile mesylate on prolactin and growth 18 155–162. hormone secretion in mice. Hormone and Metabolic Research 8 Heding LG 1972 Determination of total serum insulin (IRI) in 332–336. insulin-treated diabetic patients. Diabetologia 8 260–266. Sorenson RL, Brelje TC & Roth C 1993 Effects of steroid and Hinegardner RT 1971 An improved fluorometric assay for DNA. lactogenic hormones on islets of Langerhans: a new hypothesis for Analytical Biochemistry 39 197–201. the role of pregnancy steroids in the adaptation of islets to Holstad M & Sandler S 1993 Aminoguanidine, an inhibitor of nitric pregnancy. Endocrinology 133 2227–2234. oxide formation, fails to protect against insulitis and hyperglycemia Spangelo BL, Hall NR, Ross PC & Goldstein AL 1987 Stimulation of induced by multiple low dose streptozotocin injections in mice. in vivo antibody production and concanavalin-A-induced mouse Autoimmunity 15 311–314. spleen cell mitogenesis by prolactin. Immunopharmacology 14 11–20. Jansson L & Sandler S 1988 The influence of cyclosporin A on the Stout LE, Svensson AM & Sorenson RL 1997 Prolactin regulation of vascular permeability of the pancreatic islets and on diabetes islet-derived INS-1 cells: characteristics and immunocytochemical induced by multiple low doses of streptozotocin in the mouse. analysis of STAT5 translocation. Endocrinology 138 1592–1603. Virchows Archiv A Pathological and Anatomical Histopathology 412 Velkeniers B, Dogusan Z, Naessens F, HoogheR&Hooghe-Peters 225–230. EL 1998 Prolactin, growth hormone and the immune system in Kolb-Bachofen V, Epstein S, KieselU&KolbH1988 Low-dose humans. Cellular and Molecular Life Sciences 54 1102–1108. streptozocin-induced diabetes in mice. Electron microscopy reveals Walker SE, Allen SH & McMurray RW 1993 Prolactin and single-cell insulitis before diabetes onset. Diabetes 37 21–27. autoimmune disease. Trends in Endocrinology and Metabolism 4 Like AA & Rossini AA 1976 Streptozotocin-induced pancreatic 147–151. insulitis: new model of diabetes mellitus. Science 193 415–417. Zhu XH, Zellweger R Wichmann MW, Ayala A & Chaudry IH Neidhart M 1998 Prolactin in autoimmune diseases. Proceedings of the 1997 Effects of prolactin and metoclopramide on macrophage Society of Experimental Biology and Medicine 217 408–419. cytokine gene expression in late sepsis. Cytokine 9 437–446. Parrott RF & Goode JA 1993 Central effects of naloxone and selected opioid agonists on cortisol and prolactin secretion in non-stressed sheep. General Pharmacology 24 101–103. Received 1 March 1999 Rabinovitch A 1993 Roles of cytokines in IDDM pathogenesis and Revised manuscript received 11 June 1999 islet -cell destruction. Diabetes Reviews 2 215–233. Accepted 18 June 1999
Journal of Endocrinology (1999) 163, 229–234