Stimulation of islet cell proliferation enhances pancreatic ductal carcinogenesis in the hamster model.

Previous studies have shown that some N-nitrosobis (2-oxopropyl)amine (BOP)-induced ductal/ductular pancreatic cancers in the hamster model develop within islets and that streptozotocin (SZ) pretreatment that caused islet degeneration and atrophy inhibits pancreatic cancer induction. Hence, it appears that in this model islets play a significant role in exocrine pancreatic carcinogenesis. To examine whether stimulation of islet cell proliferation (nesidioblastosis) enhances pancreatic exocrine cancer development, we tested the effect of the pancreatic carcinogen BOP in hamsters after induction of nesidioblastosis by cellophane wrapping. Before wrapping, hamsters were treated with SZ to inhibit pancreatic tumor induction in the unwrapped pancreatic tissues. Control groups with a wrapped pancreas did not receive SZ. Six weeks after SZ treatment, all hamsters were treated with BOP (10 mg/kg body weight) weekly for 10 weeks and the experiment was terminated 38 weeks after the last BOP treatment. Many animals recovered from their diabetes at the time when BOP was injected and many more after BOP treatment. Only nine hamsters remained diabetic until the end of the experiment. Both SZ-treated and control groups developed proliferative and malignant pancreatic ductal-type lesions primarily in the wrapped area (47%) but less frequently in the larger segments of the pancreas, including the splenic lobe (34%), gastric lobe (13%), and duodenal lobe (6%). Only a few lesions developed in the unwrapped pancreatic region of nine diabetic hamsters with atrophic islets, whereas seven of these hamsters had tumors in the wrapped area. Histologically, most tumors appeared to originate from islets, many invasive carcinomas had foci of islets, and some tumor cells showed reactivity with anti-insulin. The results show that, in the BOP hamster model, islets are the site of formation of the major fraction of exocrine pancreatic cancer and that induction of nesidioblastosis enhances pancreatic carcinogenesis.     

Alloxan-induced diabetes in the mouse: Time course of pancreatie B-cell destruction as reflected in an increased islet vascular permeability

Summary The extent to which injections of the pancreatic B-cytotoxin alloxan in C57BL/Ks mice induced an increase in islet vascular permeability, and the time course of this increase, were studied. The vascular permeability was monitored by administration of the dye Monastral blue B, which is entrapped in leaky blood vessels with intact basement membranes. The islets were visualized by a freeze-thawing technique which allows identification of stained islets. Not until four hours after the alloxan injections was there an increase in islet uptake of Monastral blue B when compared with saline-treated control animals. Thereafter the islet staining increased further. The process was accompanied by gradual development of hyperglycaemia and a reduction of number of the islets identified in the pancreatic preparations. It is concluded that alloxan causes an increase in islet vascular permeability, which appears to become manifest at a later stage than the cytotoxic B-cell degeneration.This work was supported by grants from the Swedish Medical Research Council (12x-109; K86-12P-7680-01), the Swedish Diabetes Association, the Swedish Society of Medicine, the Nordic Insulin Fund, the C. Groschinsky Memorial Fund, Syskonen Svenssons Fond and the Åke Wiberg Foundation     

Early changes in islet hormone secretion in the hamster pancreatic cancer model

The diabetic state that is seen at a high frequency in association with pancreatic cancer is characterized by elevated plasma levels of several islet hormones and by marked insulin resistance. Both the diabetic state and insulin sensitivity improve after tumor removal by sub-total pancreatectomy. Impaired glucose tolerance has also been found in the hamster pancreatic cancer model, but conflicting data regarding islet function have been reported. In order to further investigate islet function and secretion during early development of pancreatic cancer, we measured the concentrations of insulin, glucagon, somatostatin, and islet amyloid polypeptide (IAPP) in plasma, pancreatic tissue, and secretin-stimulated pancreatic juice at 12 and 27 weeks after the ductal-cell-specific carcinogen, BOP had been used to induce tumors in Syrian golden hamsters. At 12 weeks after BOP, plasma glucagon levels were significantly increased. An exaggerated plasma-glucose response and concomitant hyperinsulinemia were observed at 27 but not 12 weeks after BOP. Plasma IAPP concentrations, but not glucagon or somatostatin, were elevated at 27 weeks. Tissue concentrations of IAPP were substantially reduced in BOP-treated hamsters at 27 weeks. No differences in hormone concentrations were seen in pancreatic juice from the two groups at either of the two time points investigated. The study showed that islet hormone changes accompany the early development of pancreatic tumors in the hamster pancreatic model. The hormone changes and apparent insulin resistance resemble the metabolic changes found in humans with pancreatic cancer.     

DEGRANULATION AND REGRANULATION OF PANCREATIC ISLET CELLS OF ALLOXAN DIABETIC RATS: I. HISTOCHEMICAL STUDY

The process of degranulation and regranulation of pancreatic islet cells was followed in the male rats given one intramuscular or intraperitoneal injection of alloxan (200 mg/Kg). The serum glucose showed a typical triphasic curve, reaching permanent hyperglycemia 24 hours after injection. The insulin levels of both serum and pancreas have been subnormal for the first 24 hours and then marked decrease of insulin levels was observed on the third day thereafter. In accord with subnormal pancreatic insulin levels, B granules have been present in the necrotic cells up to the first 24 hours, then complete disappearance was observed on the third day. Histochemistry of zinc was parallel to the preservation of B cells. Reactive zinc decreased from the center of the islets In 2–4 hours after alloxan, and after 24 hours zinc reaction was positive only around the periphery of the islets. Regranulated islet cells were observed from 7 days through 28 days in the presence of permanent diabetes with decreased population of B cells.     

Modification of pancreatic carcinogenesis in the hamster model. 2. The effect of partial pancreatectomy.

The effect of partial pancreatectomy (PP) on the pancreatic carcinogenicity of N-nitrosobis (2-oxopropyl)amine (BOP) was investigated in Syrian golden hamsters by subcutaneous injection of a single dose of BOP (20 mg/kg, body weight) given 30 minutes after (Group 1), 1 week after (Group 2), or 1 week before 70% PP (Group 3). Additional groups consisted of animals with PP alone (Group 4), sham operation (laparotomy) followed 30 minutes later by BOP treatment (Group 5), and BOP treatment only (Group 6). The experiment was terminated 46 weeks after BOP administration in each group. The pancreas and extrahepatic bile ducts, including the common duct and gallbladder, were examined histologically. Tumor patterns were compared in hamsters with PP and in the corresponding segments of the pancreas in BOP-treated control groups. The pancreatic cancer incidence was highest (31%) in Group 2 and lowest in Group 1 (3%), a difference that was statistically significant (P less than 0.01). Also, a statistically highly significant larger number of tumors occurred in Group 2, compared with group 1, 3, or 5 (P less than 0.0005). In a comparison of the number of carcinomas per tumor-bearing hamster, there were greater numbers of carcinomas in Group 2 (2.6 carcinomas) than in Groups 1, 3, 5, and 6 (1.0, 1.0, 1.3, and 2.6 tumors, respectively). Moreover, pancreatic tumors in Group 2 hamsters were larger (average diameter, 10 mm) than in Group 1 (4 mm), Group 3 (3.5 mm), Group 5 (4 mm), and Group 6 (average, 9mm). The incidence of extrapancreatic tumors did not vary among the PP groups but was equally lower than those in BOP-treated control groups. The data indicated BOP carcinogenesis was inhibited by surgery (regardless of whether PP was per formed) when the carcinogen was given 30 minutes after the surgery but was significantly enhanced when BOP was administered 1 week after PP. The possible reasons for these conflicting results are discussed. Morphologically all tumors were of ductular, ductal, and mixed ductular-insular patterns and most developed at the resected margins, where proliferation of islets, ducts, and ductules, but not of acinar cells, occurred. The results confirm our view that the ductal and ductular cells are the progenitor cells for BOP-induced pancreatic tumors in hamsters.     

Pancreatic uptake of [2-(14)C]alloxan

The uptake of [2-(14)C]alloxan by the pancreatic gland was investigated in control and streptozotocin-induced diabetic (STZ) rats, using both in vitro and in vivo techniques. Whether after 10 to 60 min incubation of pieces of pancreas in the presence of [2-(14)C]alloxan or 60 min to 24 h after intravenous injection of [2-(14)C]alloxan to control and insulin-treated STZ rats, the radioactive content of the pancreas (dpm/mg wet weight) only represented, in the STZ rats, about two thirds of the reference value found in control animals. These findings indicate that insulin-producing islet B-cells participate to a sizeable extent to the overall uptake of [2-(14)C]- alloxan by the whole pancreatic gland, despite the fact that they account for no more than about one percent of the total pancreas mass. Hence, it should be possible to preferentially label the endocrine moiety of the pancreas, in the perspective of its imaging and quantification by a non-invasive procedure, by use of a suitable radiolabelled molecule selectively taken up by islet, as distinct from acinar, pancreatic cells.     

Modification of pancreatic carcinogenesis in the hamster model. 7. Inhibitory effect of bethanechol chloride

Experiments were designed to investigate in the hamster model the effect on pancreatic carcinogenesis of bethanechol chloride (BC), which is known to increase pancreatic protein synthesis in rats. Hamsters received a single (15 mg/kg body weight) dose of BC either before, simultaneously with, or after a single dose of N-nitrosobis(2-oxopropyl)amine (BOP; 20 mg/kg body weight). A second group was treated daily with BC (7.5 mg/kg body weight) for 24 weeks, following BOP. The control groups consisted of animals treated with BOP only, with BC only, and with solvent only. The surviving hamsters were killed 46 weeks after BOP treatment. BC, whether given before, simultaneously with, or after BOP, significantly reduced the incidence of pancreatic ductal/ductular carcinomas. The multiplicity, size, and latency of carcinomas were also affected by BC. A more pronounced inhibition of cancer induction occurred in the group treated daily with BC after BOP. The possible mechanisms involved in the inhibitory action of BC on pancreatic carcinogenesis are discussed.     

In vitro induction of giant cell tumors from cultured hamster islets treated with N-Nitrosobis(2-Oxopropyl)amine

Giant cell carcinoma of the pancreas is a rare tumor. Its histogenesis is still controversial. In a Syrian hamster pancreatic cancer model, tumors similar to human giant cell carcinomas have been induced at an extremely low rate of incidence and after the use of high doses of pancreatic carcinogens. Thus far no tumors of giant cell type have been induced by the in vitro treatment of hamster pancreatic ductal cells with the potent pancreatic carcinogen N-nitrosobis(2-oxopropyl)amine (BOP). In the present study we report the induction of giant cell carcinoma from hamster islets treated with BOP in vitro. The results suggest that in hamsters some component of islet cells, probably stem cells, are the origin of giant cell carcinoma.     

Diabetogenic effects of n-nitrosomethylurea in the chinese hamster.

To evaluate the role of the glucose residue of the diabetogenic substance streptozotocin, the effect of its aglucone derivate N-nitrosomethylurea was tested in Chinese hamsters. At a dose of 50 mg/kg body weight of N-nitrosomethylurea a triphasic blood glucose curve was recorded with an initial hyperglycaemic peak after 3 hours followed by a decrease at 6 hours. After 24 hours and during the following days the values were moderately elevated. There was a high mortality in the diabetic animals, about 80 per cent of them dying within one week. The approximate L.D. 50 of N-nitrosomethylurea injected intraperitoneally in non-fasting adult animals was about 125 mg/kg body weight at an observation time of 48 hours. On light microscopy, degenerative changes with nuclear pyknosis were seen after 3 hours in the pancreatic islet cells, followed by cellular disintegration. Both beta-, alpha2- and alpha1-cells were obviously affected. Pretreatment with 500 mg nicotinamide/kg body weight given intraperitoneally 10 minutes before the injection of N-nitrosomethylurea inhibited and hyperglycaemia and seemed to prevent the injurious effects of N-nitrosomethylurea on the islet cells. The results show that the glucose residue of the streptozotocin molecule is not necessary for the induction of diabetes in Chinese hamsters, but it seems to increase the selectivity of the toxic effects for the islet cells. This is a clear advantage in studies of experimental diabetes, especially when longer observation periods are desired.     

Chemically induced (streptozotocin-alloxan) diabetes mellitus in the dog. Biochemical and ultrastructural studies

Adult beagle dogs were infused intravenously with alloxan (50 mg/kg) and streptozotocin (30 mg/kg) in order to investigate sequential changes in plasma glucose, insulin, glucagon, and cortisol. These biochemical findings were correlated with ultrastructural alterations in pancreatic islets. Following infusion, the dogs became hyperglycemic by 2 hours, severely hypoglycemic by 6 to 14 hours, and permanently hyperglycemic by 24 hours. Plasma immunoreactive insulin increased sharply from 6 to 10 hours, then declined to nearly undetectable levels by 24 hours. Ultrastructurally, beta cells at 2 hours had clumped nuclear chromatin, vacuolated mitochondria, and dilated individual profiles of endoplasmic reticulum. Secretory granules appeared swollen but retained their internal cores. Degeneration of beta cells was severe after 10 hours, and plasma membranes of beta cells were disrupted by 24 hours. The cytoplasmic area of adjacent beta cells coalesced and had accumulations of membranous debris, lipofuscin, and autophagic vacuoles. Alpha and delta cells appeared to be unaffected. Plasma glucagon levels decreased markedly at 10 hours and were related reciprocally to changes in plasma insulin. Pancreatic islets in dogs with chronic (16 months) diabetes were small and composed primarily of granulated alpha and delta cells. Poorly granulated beta cells with degenerative changes were present in an occasional islet. The results of this investigation demonstrated that the combined administration of a single dose of alloxan and streptozotocin selectively destroyed the beta cells, while alpha and delta cells of the islets of Langerhans remained unaltered. Pathologic evidence of toxicity was not present in other organs. Chemically induced diabetes mellitus in dogs is a reproducible animal model that should prove useful in studies requiring repeated experimental manipulations or sampling of biologic fluids in order to evaluate the long-term effects of different routes of delivery or preparations of insulin to control the persistent hyperglycemia.     

State of the pancreatic islets,fetal weight,and glucose tolerance test in offspring of male rats with alloxan diabetes

The body weight and morphometric parameters of the state of the pancreatic insular apparatus (the relative percentage of islet tissue, of A- and B- cells, the number of islets) in fetuses and the adult offspring of healthy female rats and male rats with alloxan diabetes were studied. No change in these parameters was found, indicating that under these conditions the genetic systems of the males are undamaged and that these lesions are not transmitted to the progeny. On the basis of these findings and of the importance of diabetes in the female in the development of insular deficiency in the progeny demonstrated previously, the role of the environment in which the fetus develops in the formation of the diabetes is emphasized.     

Cytotoxic effects of alloxan treatment in vitro on monolayer cultures of neonatal rat pancreas

Monolayer cultures of neonatal rat pancreas were exposed briefly (5 min) to alloxan and were then maintained for several days (up to 14 days) to examine the long-term effects of alloxan exposure on B cells. Alloxan (0-10 mM) caused a dose-dependent reduction in rates of insulin release during the first 24 h after treatment. This reduction persisted throughout the period of observation. Rates of glucagon release were unaltered by all concentrations of alloxan tested. Spontaneously decomposed alloxan had little or no effect on rates of insulin release. The simultaneous presence of high concentrations of glucose during exposure of cells to alloxan exerted a rapid protective effect against alloxan toxicity that was both glucose and alloxan dose dependent. Alloxan treatment of pancreatic monolayer cultures may allow for a comparison of those actions of alloxan that are directly on and specific for islet B cells. Furthermore, alloxan-treated cultures may represent a unique in vitro system for studying the physiology and biochemistry of other islet cell types in cultures in which B cells and insulin are reduced, as well as for studying glucose interactions with the B cell.     

Further characterization of carcinogen-induced hepatocytelike cells in hamster pancreas.

Regenerating pancreatic cells of the Syrian hamster treated at the peak of S phase with the pancreatic carcinogen N-nitrosobis(2-oxopropyl)amine (BOP) are converted into stable cells wtih morphologic and functional characteristics that are strikingly similar to those of differentiated hepatocytes. In this article the authors further document their hepatocytelike nature. Seventy-two hours after subtotal hepatectomy, pancreatic hepatocytelike cells responded with an 8-fold increase in labeled nuclei (105.8 +/- 4.04/1000 cells) which had incorporated 3H-thymidine and a 5-fold increased mitotic index (3.8 +/- 1.5 mitoses/1000 cells), as compared with similar cells in the pancreas of control animals that had undergone sham operations. Chronic administration of phenobarbital induced a 31-fold increase in the level of aryl hydrocarbon hydroxylase (AHH) in pancreas containing such cells, as compared with normal control pancreases, and caused marked proliferation of smooth endoplasmic reticulum (SER). These cells also showed an enhanced capacity for the accumulation of iron during acute iron excess, as compared with adjacent acinar cells. Collectively, these findings support the view that carcinogen-induced cells in pancreas bear a close functional resemblance to hepatocytes.     

The fate of intraportally transplanted islets in diabetic rats. A morphologic and immunohistochemical study.

Streptozotocin-induced diabetes in the rat can be reversed by the transplantation of isogenic islets of Langerhans from neonatal donors. We studied the morphology of intraportally transplanted islets with the aid of the immunoperoxidase staining technique to identify insulin-, glucagon-, somatostatin-, and pancreatic polypeptide-containing cells at 24 hours, 48 hours, 1 week, 2 weeks, 4 weeks, 39 weeks, and 65 weeks after transplant. Embolized pancreatic tissue, composed of approximately 80% acini and 20% islets, is initially distributed throughout the liver mainly to terminal branches of the portal system. Endothelialization and organization occur rapidly with the smaller fragments and within the first 4 weeks for larger thrombi. Exocrine pancreatic elements largely disappear as islet cells move into the hepatic lobules from the portal spaces. At 65 weeks after transplant, all islet cell types can be identified within large complex islet structures. The results of this study establish the survival and continued function of all known rat pancreatic islet cell types long after transplantation and support the theory that islet transplantation may represent the most physiologic replacement of hormonal deficiencies in the diabetic recipient.     

Glomerular basement membrane thickness following islet transplantation in the diabetic rat.

Glomerular basement membrane (GBM) thickness was measured in diabetic rats prior to and following islet transplantation. Over the course of the experiment (from 7 to 13 months of diabetes) GBM thickness in diabetic animals exceeded that of littermate controls. After 7 months of diabetes a group of animals received successful intraportal transplants of neonatal pancreatic tissue. GBM thickness at 2 and 6 months following islet transplantation matched the thickness in nontransplanted diabetic rats and exceeded that in control animals. Failure to reverse GBM thickening in diabetic rats following islet transplantation may be due to very slow rates of GBM tunover in the rat. Previous work has demonstrated normalization of urinary albumin excretion after islet transplantation, suggesting that GBM thickening, per se, is not a significant factor causing albuminuria in rats with longstanding diabetes.     

Modification of pancreatic carcinogenesis in the hamster model. 6. The effect of ductal ligation and excision

The effect of ligation and excision of the pancreatic duct in pancreatic carcinogenesis was examined in the hamster model. Animals were treated with a single dose (20 mg/kg body weight) of N-nitrosobis(2-oxopropyl)amine (BOP) either immediately (Group 1) or on Days 1 (Group 2), 3 (Group 3) or 7 (Group 4) after ligation and excision of the duct of the splenic lobe. Group 5 received BOP shortly after laparoscopy, and Group 6 consisted of BOP-treated controls. All hamsters were killed 46 weeks after BOP treatment. The results showed that despite advanced atrophy of the splenic lobe distal to the excised duct in Groups 1-4, some hamsters in Groups 2, 3, and 4 showed hyperplasia, dysplasia, and increased mitotic activities of ductal and ductular cells. However, carcinomas in the duct-excised atrophic lobe were found only in Groups 1-3. These data indicate that BOP carcinogenesis is mediated through blood circulation, and that cancer development is not inhibited in the duct-excised lobe for up to 3 days after surgery. However, in the entire pancreas, a significant reduction in tumor incidence was seen when the carcinogen was given immediately, or to a lesser extent, 1 day after surgery, regardless of whether or not excision was made. On the contrary, BOP, when given 3 and 7 days after duct excision, enhanced tumor development in the nonexcised (intact) pancreas, compared with other test groups and with BOP controls. Both inhibition and enhancement seemed due to a proportional decrease and increase, respectively, of BOP-responsive cells throughout the intact pancreas.     

Insulinomas derived from hyperplastic intra-hepatic islet transplants.

Insulinomas were induced in a new animal model by transplanting a low number of isologous pancreatic islets via the portal vein into the livers of 66 streptozotocin-induced diabetic rats. In contrast to high-number islet transplantation, which restored normoglycemia in 25 control animals, the low-number islet transplantation was followed by persisting hyperglycemia for at least 13 months. Hyperplasia of islet cells developed in the transplanted islets as a consequence of hyperglycemia, which for the beta cells is not only a secretory but also a proliferative stimulus. Six of thirty-three animals between the 18th and the 24th month after islet transplantation changed from hyperglycemia to severe hypoglycemia, due to insulinomas that had developed in the liver from the transplanted islets. In contrast to other animal models, insulinoma development in this new model does not result from DNA damage by chemicals or radiation or from the expression of transgenes, but starts from apparently normal islets prepared from untreated isologous donors, which are exposed to an imbalance in glucose metabolism. The persistent proliferative stimulus and the metabolic alterations caused by the longstanding hyperglycemia seem to be the most relevant oncogenic factors in this model.     

Dissociation between pancreatic islet blood flow and insulin release in the rat

The blood flow to the pancreatic islets was estimated with the aid of microspheres in fed or starved (72 h) rats. The total pancreatic blood flow (PBF) in fed animals was 0.55 +/- 0.04 ml X min-1 X g pancreas and in the starved animals 0.30 +/- 0.04 ml X min-1 X g pancreas (P less than 0.001), and the corresponding islet blood flow (IBF) 82.0 +/- 12.4 and 50.5 +/- 9.7 microliter min-1 X g pancreas respectively (P greater than 0.05). Intraperitoneal injection of 2 ml of a 30% glucose solution caused a marked increase in IBF in both fed (P less than 0.05) and starved (P less than 0.01) animals to approximately the same level. The circulating insulin concentration remained unaffected by glucose in the starved rats but increased (P less than 0.001) in the fed rats, indicating that insulin release does not necessarily rise in parallel with an elevated IBF. Intraperitoneal injection of 2 ml of a 30% solution of mannoheptulose, an inhibitor of islet glucose metabolism, decreased the serum insulin concentrations although the serum glucose concentrations rose significantly in both fed (P less than 0.001) and starved (P less than 0.001) animals. This treatment, however, caused both IBF and PBF to increase significantly in both groups. The data support the view that islet blood flow is not necessarily related to the metabolic status of the islet cells or to the insulin release.     

Pancreatic lesions induced in rabbits and guinea-pigs with pancreatic antigens.

Rabbits and guinea-pigs were immunized with various pancreatic antigens in Freund's adjuvant. Rabbits received unfractionated bovine insulin and the "A" component and "single peak" insulin separated from it by gel-filtration. All produced antibodies capable of reacting with porcine insulin but none were found to have pancreatic lesions when killed up to 6 weeks after initial injection. Guinea-pigs immunized with bovine "A" component developed pancreatic peri-ductulitis which appeared most frequently (10/20) in animals killed 30 days after a single injection and less frequently in animals killed after 60 (4/10) and 90(1/10) days. Similar lesions were found in only a small proportion of control animals (2/23) or of guinea-pigs immunized with single peak bovine insulin (3/22). Guinea-pigs immunized with homogenates of homologous and heterologous islets of Langerhans developed signs of peri-ductulitis in a high proportion of animals killed up to about 60 days after first injection (18/26). None of these animals exhibited clearly defined signs of diabetes mellitus and the incidence of induced lesions could not be correlated with levels of circulating insulin-binding antibodies.     

An ultrastructural study of the effect of a diabetogenic dose of alloxan on the secretory ameloblasts of the rat incisor

The effect of a single diabetogenic dose of alloxan on the ameloblasts of enamel secretion was investigated in rat incisor teeth prior to the onset of diabetes mellitus. This was compared with tissue from animals that were sacrificed at the onset of diabetes, and with tissue from animals that had been diabetic for 1 month. Male Sprague-Dawley rats were given a single subcutaneous injection of alloxan at a dose of 150 mg/kg body weight; and the animals were sacrificed at 45 minutes, 2 hours, 24 hours, 48 hours and 29 days after injection. At the electron microscope level the following changes were observed. There was an early accumulation of secretion granules in the vicinity of the Golgi apparatus and within the proximal portion of Tomes' process. These accumulations were present at the onset of diabetes, 24 hours later. At 2 hours after injection a space appeared between the plasma membrane around Tomes' process and the interrod material. This space widened with time, and at 24 hours after injection it became continuous with enlarged intercellular spaces between the proximal portions of Tomes' processes. The widening of the intercellular space was restricted to the area above the distal cell web. At the onset of diabetes this compartment of the cells was similar to that of the control samples. Extracellular material which appeared to be the secretory product of the ameloblast was observed at two sites. On one occasion this material was seen in the wide intercellular spaces between the proximal portions of Tomes' processes 24 hours after injection. It was also seen at the onset of diabetes in the intercellular space at the level of the Golgi apparatus. The changes in the animals that had been diabetic for 1 month were a scarcity of secretion granules within Tomes' processes and an abnormal accumulation of secretion granules within the supranuclear and infranuclear compartents. This study has shown that the toxic effect of alloxan persists up to the time of onset of diabetes mellitus. However, since different morphological changes were observed during diabetes mellitus, it is suggested that the changes caused by diabetes are a separate entity, initially superimposed on and later replacing the acute, toxic effect of the drug.