Proc. NatI. Acad. Sci. USA Vol. 74, No. 6, pp. 2485-2489, June 1977 "I ", .4-4" I- Cell Biology ,- -i,' Studies of streptozotocin-induced insulitis and diabetes (pancreatic beta cells/islets of Langerhans/alloxan diabetes/type C virus induction/cell-mediated reaction) ALDO A. ROSSINI*, ARTHUR A. LIKEt, WILLIAM L. CHICK*, MICHAEL C. APPELt, AND GEORGE F. CAHILL, JR. * * Joslin Research Laboratory, One Joslin Place, Boston, Massachusetts 02215 and Harvard Medical School and Peter Bent Brigham Hospital, Boston, Massachusetts 02215; and t Department of Pathology, University of Massachusetts Medical School, Worcester, Massachusetts 01605 Communicated by Eugene Braunwald, April 4, 1977 ABSTRACT Multiple small injections of streptozotocin was a subsequent decrease in inflammation within the re- produce a delayed, progressive increase in plasma-glucose in maining islets which were small and composed almost exclu- mice within 5-6 days after the injections, in association with sively of non-beta cells. Blood glucose values remained in the pronounced insulitis and induction of type C viruses within beta cells. Multiple subdiabetogenic doses of streptozotocin in rats diabetic range and correlated well with the pathologic changes and multiple injections of another beta cell toxin, alloxan, in within the islets. Noteworthy also was the presence of numerous mice did not induce insulitis although hyperglycemia followed type C viruses within the surviving beta cells of animals studied the injection of larger quantities of both agents. In mice, the within 6 days of the last SZ injection. The delayed development prior injection of 3-O-methyl-D-glucose (3-OMG) or nicotin- of the insulitis and the nature of the inflammatory infiltrate amide attenuated the diabetic syndrome produced by strepto- zotocin; however, 3-OMG was more protective. Rabbit anti- appeared to suggest a cell-mediated immune reaction. The role mouse lymphocyte serum, alone, provided partial protection of the increased number of type C viruses within the beta cell but, when given together with either 3-OMG or nicotinamide, remains to be clarified. effectively prevented the streptozotocin-induced diabetic syn- In this paper we report additional studies of the streptozo- drome. Cessation of these preventive treatments was followed tocin-induced insulitis model of diabetes. Various substances by the appearance of insulitis and diabetes. These findings that alter or attenuate the insulitis were studied, including 3- suggest that multiple injections of streptozotocin induce, in susceptible hosts, the triad of direct beta cell cytotoxicity, virus O-methyl-D-glucose (3-OMG) and nicotinamide which have induction within beta cells, and cell-mediated autoimmune been shown to protect against the single-injection method -of reaction. These factors, acting separately or in concert, appear inducing diabetes (9). Antilymphocyte serum (ALS), which has to induce a destructive insulitis and severe diabetes. The relative been used to inhibit cell-mediated graft rejection (11-13), was importance of each component and the factors governing host used to determine indirectly whether the SZ injections trigger susceptibility remain to be clarified. a cell-mediated autoimmune reaction directed against the pancreatic beta cells. Furthermore, the specificity of the re- Streptozotocin [SZ; 2-deoxy-2(3-methyl-3-nitrosoureido)-D- sponse to SZ was evaluated by comparison with another beta- glucopyranose] is a broad-spectrum antibiotic with oncolytic, cytotoxic diabetogenic agent, alloxan. Finally, experiments oncogenic, and diabetogenic properties (1-4). The diabetogenic the of rats to injections of SZ were action is mediated by selective destruction of pancreatic beta evaluating response multiple cells and has been widely utilized as a method for inducing also carried out to determine if insulitis and hyperglycemia diabetes mellitus in experimental animals and for treatment could be induced in another species. of malignant beta cell tumors and other neoplasms in humans. To produce diabetes, SZ is conventionally administered as a MATERIALS AND METHODS single injection. SZ is cleared from the bloodstream rapidly (serum half-life, 15 min) (5); beta cell necrosis can be detected Animals. Male rats (Charles River strain of Sprague-Dawley) by electron microscopy within hours after SZ injection (6-8). weighing 190-200 g and male mice (Charles River CD-1 strain) Elevated blood glucose levels are demonstrable within 1-2 days, weighing 35-40 g were obtained from Charles River Breeding and dissolution and phagocytosis of necrotic cells are observed Company, Wilmington, MA. The rats were given free access histologically after 3 days (3, 6). In rats and mice, the admin- to water and Purina Rat Chow; the mice were given Old istration of a single subdiabetogenic dose produces only mild Guilford Mouse Pellets (96W). histologic alterations without evidence of significant hyper- Toxic Compounds. Streptozotocin (U9889, lots 11837- glycemia when compared with buffer-injected control animals GGS-22B and 1676-RCH-112) was kindly supplied by W. E. (9, 10). Dulin of the Upjohn Co., Kalamazoo, MI. The crystalline SZ We have recently reported (10) that the administration of was assayed to be approximately 90% a anomer based on optical multiple (five) subdiabetogenic doses of SZ to Charles River rotation data. SZ was dissolved in a citrate buffer, pH 4.5, and Laboratory (CD-1) mice, either intravenously or intraperito- injected intravenously or intraperitoneally within 15 min of neally, produced a delayed but progressive increase in blood dissolution; Alloxan monohydrate (Eastman Co., Rochester, NY) glucose with levels of 350450 mg/100 ml within 5-6 days after was dissolved in 0.9% NaCl just prior to use and was injected the last injection. Morphologically, the pancreatic islets revealed intravenously in a volume of 0.1 ml. An additional 0.2 ml of pronounced insulitis with infiltrating lymphocytes and mac- 0.9% NaCI was given to flush the solution through the tub- rophages, architectural distortion, and beta cell necrosis. There ing. Histology and Plasma Glucose Determinations. For light Abbreviations: SZ, 2-deoxy-2(3-methyl-3-nitrosoureido)-D-glucopy- microscopy, the pancreata were fixed in Bouin's solution, and ranose; ALS, antilymphocyte serum; 3-OMG, 3-0-methyl-D-glu- sections were stained with hematoxlin and eosin and aldehyde cose. fuchsin. For electron microscopy, tissues were fixed in a solution 2485 24862Proc.Cell Biology Rossini ct al. N'aitl. Acad. Sci. 5S-A 7Ti (1977; 700* 700} 'A ) 6 f..! 61 -1ca)e3: - - -.. 3 fs - 1 6toJ00 .>3f r! r- Iv 103 a.0 500- /24) ;74 Z7 ue 41>0 ,AL111) v' 4 L 400 w.- ''24' X p '4 9/401 -D 30 110 / ,,, i -.300, - .'4""12-,1 2 5? 724'.i 9!' 16GX :-- '6G 200 :1 cr Q *i - 1 6 .11 _i _r i) 61t 6' ' C~---G-sA (O. ~ .lo 0 C) 1 3 4 5 / 10 111 1314 1 4 A)'1?]4'4 Or) 71 9 1 C 1 1 21 4 1 53 4 [DAYS WEEKS [) >\v E E- KS FIG. 1. Plasma glucose (mean ± SEM) levels in CD-1 mice after: FIG. 2. Plasma glucose levels (mean ± SEM) in CD-1 mice after: 5 daily intraperitoneal injections of SZ (40 mg/kg) (0); 10 daily in- 5 intraperitoneal injections of SZ (40 mg/kg) (0); 10 intraperitoneal traperitoneal injections of ALS (0.5 ml) plus SZ (0); 10 intraperitoneal injections of ALS (0.5 ml) plus SZ (0); 10 intraperitoneal injections injections of 3-OMG (0.22 mmol per mouse) plus SZ (A); or 10 in- of nicotinamide (0.22 mmol per mouse) plus SZ (A); or 10 intraperi- traperitoneal injections of 3-OMG plus ALS plus SZ (A&). Numbers toneal injections of ALS plus nicotinamide plus SZ (A&). Numbers in in parentheses are numbers of determinations. parentheses are numbers of determinations. of paraformaldehyde and glutaraldehyde as reported (14). Blood samples for glucose determination were collected from amide, or ALS was continued for 5 days after completion of the fed animals into heparin-treated pipettes and assayed as de- SZ injections. scribed previously (10). Fig. 1 illustrates the results of glucose determinations over Other Substances. Nicotinamide and 3-OMG were obtained a 4-week period during which SZ was given for 5 days alone or from the Sigma Chemical Co., St. Louis, MO. The 3-OMG was in combination with either 3-OMG (0.22 mmol per mouse) or dissolved in 0.9% NaCi at least 2 hr prior to injection to ensure ALS (0.5 ml per mouse), or both 3-OMG and ALS. In agreement mutarotation of the anomers to equilibrium. The solution was with our previous findings (10), glucose levels in SZ-injected injected in a volume of 0.1 ml (0.22 mmol per mouse). Nico- mice were elevated on the fifth day of injections and continued tinamide was prepared in a similar manner, and 0.22 mmol per to increase progressively in the subsequent 3-week period. The mouse was injected intraperitoneally in a volume of 0.1 ml. animals receiving the combination of 3-OMG and SZ evidenced Rabbit anti-mouse lymphocyte serum (ALS) and normal no increase in glucose for the initial 12 days; however, by the (nonimmune) rabbit serum were obtained from Microbiological 15th day, 5 days after the cessation of 3-OMG injections, glucose Associates, Bethesda, MD. The ALS preparations were tested elevations were observed and persisted for the following 2 by hemagglutination, cytotoxicity, and skin allograft prolon- weeks. It is noteworthy that plasma glucose levels in this group gation:t lot 15172, cytotoxic at 1:3200, caused hemagglutination were significantly lower than in those mice receiving SZ alone. at 1:128, and resulted in skin graft prolongation of 30.3 + 1.8 Among the animals sacrificed on the 12th day of the experi- days; lot 15096, cytotoxic at 1:6400, caused hemagglutination ment, histologic examination revealed only an occasional round at 1:64, and resulted in skin graft prolongation of 30.1 i 3.3 cell infiltrating or surrounding the otherwise normal-appearing days.