Acinar Cell Necrosis Research Articles

Failure of a Potent Cholecystokinin Antagonist to Protect Against Diet-Induced Pancreatitis in Mice

The effects of a potent cholecystokinin (CCK) receptor antagonist, L-364,718, on two forms of experimental acute pancreatitis in mice were evaluated. The antagonist prevented the hyperamylasemia, pancreatic edema, and acinar cell vacuolization that followed administration of a supramaximally stimulating dose of the cholecystokinin analogue cerulein. In contrast, the same dose of L-364,718 (1 mg/kg/6 h) and an even higher dose (10 mg/kg/6 h) failed to prevent the hyperamylasemia, acinar cell necrosis, and mortality that followed administration of a choline-deficient ethionine-supplemented diet. These observations are at variance with those previously reported to follow administration of the relatively weak cholecystokinin antagonist proglumide (Niederau C et al. J Clin Invest 1986;78:1056-63). The observations reported in this communication suggest that cholecystokinin does not play an important role in diet-induced pancreatitis and that CCK receptor antagonists are unlikely to be of benefit in the treatment of clinical acute pancreatitis.

Morphological, biochemical and autoradiographic studies of pancreatic acinar cell necrosis and its regeneration induced by 4-hydroxyaminoquinoline-1-oxide in rats

A single intravenous injection of 4-hydroxyaminoquinoline-1-oxide (4-HAQO) induced a dose-dependent diffuse pancreatic acinar cell necrosis within 48 hours after the administration in rats. Regeneration following the necrosis ensued by 72 hours. The serum amylase levels were markedly elevated within 24 hours, which paralleled a decrease of digestive enzyme contents in the pancreas. The change of these biochemical parameters corresponded well with the morphological change of the pancreas. Autoradiographic studies with 3H-thymidine revealed that the labeling index of the regenerating acinar cells was considerably higher, reaching a peak value of nearly 18.2% at 72 hours after injection of 4-HAQO and it still remained high in 2.67% even at 168 hours after injection.

Experimental Encephalomyocarditis Virus Infection in Mongolian Gerbils (Meriones unguiculatus)

Two strains of Mongolian gerbils (Meriones unguiculatus), Tumble Brook (TUM) and Japan Medical Science (JMS), were intraperitoneally inoculated with the D variant of encephalomyocarditis virus (EMC-D) and killed 3 days later. Mortality was significantly higher in females than in males. Evidence of viral replication was detected in the heart of both strains and in the pancreas of the TUM strain. Histopathological alterations were found in the heart and pancreas. Heart lesions involved foci of necrosis with inflammatory cell infiltration and calcification in both strains. Pancreatic lesions were restricted to the exocrine glands; islets of Langerhans were rarely and secondarily involved in the extensive destruction of exocrine glands. Severe acinar cell necrosis with marked inflammatory edema was conspicuous in TUM, whereas only slight acinar cell involvement was detected in JMS gerbils. Immunoperoxidase staining showed viral antigens in intracytoplasmic vacuoles in damaged acinar cells.

Ductal vasculature in the atrophic pancreas: evidence for autonomous ductal angioarchitecture.

Although it is known that ducts remain intact in the atrophied pancreata of rats rendered dietarily copper-deficient, no information is available that characterizes the ductal angioarchitecture of the experimentally altered gland. Furthermore, since the vascular relationship between pancreatic ducts and the parenchyma of the normal gland has not been well defined, this atrophy model was used to examine blood flow to ducts in the atrophied gland where acinar tissue undergoes selective necrosis. Scanning electron microscopy (SEM) of pancreatic corrosion casts produced from Mercox-injected rats was used. SEM was compared with light microscopic study of stain-injected cleared pancreata and correlated with histological and ultrastructural studies. The results indicate that duct cells and ductal vasculature remain unaffected by surrounding acinar cell necrosis and glandular lipomatous atrophy. Furthermore, the results support a largely autonomous blood supply for the ductal system by arterial blood that flows directly from major extralobular arteries of the pancreas or their interlobular branches. In the same manner, venous blood draining the ductal system returns to major extralobular veins or by way of their interlobular branches. No evidence could be established for manifest acinoductal or ductoacinar portal systems. There were, however, places in the route of both ductal and acinar vessels where the vessels were found to anastomose. These places were found near the source of major arteries or veins that supply and drain confluent acinar and ductal tissues. It is concluded that, while some acinoductal or ductoacinar exchange may take place, the role of these putative portal systems would appear less significant than the direct route of blood to and from pancreatic ducts.

Poliomyelomalacia, Pancreatic Necrosis, and Cerebellar Malacia in Turkey Poults

Two-to-5-week-old turkey poults from three large Minnesota flocks exhibited ataxia, flaccid paralysis, and up to 5% mortality as unexpected death. The major post-mortem finding was cerebellar hemorrhage and softening detected in 22 of 89 clinically affected poults. Histologic findings were severe focal or multifocal poliomyelomalacia in the lumbosacral intumescentia of the spinal cord, cerebellar malacia, and single-cell or multifocal coagulative necrosis of pancreatic acinar cells. Thirty of 32 clinically affected poults examined had microscopic spinal cord lesions, 12 of 48 had cerebellar lesions, and 26 of 47 had pancreatic lesions. Gross and microscopic cerebellar lesions resembled those of vitamin E deficiency in chicks. Hepatic selenium levels were approximately twice normal expected levels for poults.

Experimental T-2 toxicosis in swine: III. Morphologic changes following intravascular administration of T-2 toxin

The gross and microscopic changes in swine following a single intravascular (iv) dose of T-2 toxin are described and evaluated quantitatively. T-2 toxin, in 70% ethanol, was given iv at 0 (5 pigs), 0.6 (5 pigs), 1.2 (1 pig), 4.8 (5 pigs), or 5.4 (2 pigs) mg/kg to 40 to 60 kg female crossbred pigs. The 4.8 and 5.4 mg/kg group pigs died between 5 and 10.5 hr after treatment, while the 0, 0.6, and 1.2 mg/kg pigs were killed at 24, 24, and 12 hrs after treatment, respectively. Morphologic examination was performed at the gross and light microscopic levels. In addition, a quantitative evaluation of microscopic changes present in lymphoid tissues and intestinal tract was performed using a semiquantitative scoring system. Gross lesions in the T-2-treated pigs consisted of edema, congestion, and hemorrhage of the lymph nodes and pancreas; congestion and hemorrhage of the gastrointestinal mucosa, subendocardium, adrenal gland, and meninges; and edema of the gall bladder. Histologic examination confirmed the gross observations. Additional microscopic lesions included widespread degeneration and necrosis of the lymphoid tissues as well as of the surface and crypt epithelium of the gastrointestinal mucosa; mild scattered necrosis of pancreatic acinar cells, myocardium, bone marrow cells, adrenal cortical cells, and tubular epithelium of renal medulla; and mild interstitial pneumonia. A dose-dependent increase in lesion severity was observed except for the pancreatic lesion which was slightly more apparent in the pigs from the 0.6 mg/kg group. These findings indicate that (1) T-2 toxin-induced lesions in the lymphoid tissues and gastrointestinal tract of pigs are similar to those of other species, (2) the pancreas and heart should be considered as additional target organs in the pig, and (3) both rapidly dividing cells and those with little or no turnover are damaged by T-2 toxin.

The morphological changes of the pancreas in hypovolemic shock and the effect of pretreatment with steroids

Previous reports from this laboratory have described the effect of hypovolemic shock of varying duration on the exocrine function of the pancreas, and the ability of steroids to reverse the inhibition of secretion observed. This report is a study of pancreatic morphology in prolonged hypovolemia, as well as the effect of steroid pretreatment on the pathology observed. Twelve mongrel dogs were divided into two groups, one with and one without steroid pretreatment. The animals were bled until at least 30-35% of their blood had been withdrawn, or until mean blood pressure (BP) dropped to around 60 mmHg. When BP dropped to 80-90% mmHg, six animals received intravenous (i.v.) prednisolone in 50 cc of 0.9% NaCl, at the rate of 10 cc/min, for a dosage of 30 mg/kg. Of the twelve dogs, six were exposed to 120 min of hypovolemia, and six to 180 min. At the end of the observed hypovolemic period, the blood withdrawn was quickly retransfused. After an hour of recovery, the animals were killed and their pancreases removed. The pancreases were weighed, fixed and examined by light microscopy. Three pancreases were obtained as control from animals not subjected to shock or steroids. The steroid-treated animals displayed an insignificant (5%) increase in pancreatic weight following prolonged hypovolemia. Untreated canine pancreases, on the other hand, showed a significant weight gain (37%, P less than 0.001) after 3 h of hypovolemia. Microscopically, the untreated group revealed marked interstitial edema, hemorrhage and inflammation, as well as focal acinar cell necrosis and fat necrosis, while in the steroid group only very mild edema and inflammatory changes were seen. This study indicates a beneficial effect of steroids on the morphological changes seen in shock-induced pancreatitis in canines and a possible use in the therapy of acute pancreatitis in man.

Experimental T-2 Toxicosis in Swine

The gross and microscopic changes in swine following a single intravascular (iv) dose of T-2 toxin are described and evaluated quantitatively. T-2 toxin, in 70% ethanol, was given iv at 0 (5 pigs), 0.6 (5 pigs), 1.2 (1 pig), 4.8 (5 pigs), or 5.4 (2 pigs) mg/kg to 40 to 60 kg female crossbred pigs. The 4.8 and 5.4 mg/kg group pigs died between 5 and 10.5 hr after treatment, while the 0, 0.6, and 1.2 mg/kg pigs were killed at 24, 24, and 12 hrs after treatment, respectively. Morphologic examination was performed at the gross and light microscopic levels. In addition, a quantitative evaluation of microscopic changes present in lymphoid tissues and intestinal tract was performed using a semiquantitative scoring system. Gross lesions in the T-2-treated pigs consisted of edema, congestion, and hemorrhage of the lymph nodes and pancreas; congestion and hemorrhage of the gastrointestinal mucosa, subendocardium, adrenal gland, and meninges; and edema of the gall bladder. Histologic examination confirmed the gross observations. Additional microscopic lesions included widespread degeneration and necrosis of the lymphoid tissues as well as of the surface and crypt epithelium of the gastrointestinal mucosa; mild scattered necrosis of pancreatic acinar cells, myocardium, bone marrow cells, adrenal cortical cells, and tubular epithelium of renal medulla; and mild interstitial pneumonia. A dose-dependent increase in lesion severity was observed except for the pancreatic lesion which was slightly more apparent in the pigs from the 0.6 mg/kg group. These findings indicate that T-2 toxin-induced lesions in the lymphoid tissues and gastrointestinal tract of pigs are similar to those of other species the pancreas and heart should be considered as additional target organs in the pig, and both rapidly dividing cells and those with little or no turnover are damaged by T-2 toxin.

Pancreatic acinar cell regeneration following copper deficiency-induced pancreatic necrosis.

Rats maintained on a diet deficient in copper for up to 8-10 weeks exhibit marked necrosis and depletion of acinar cells in the pancreas. When these animals are subsequently fed a copper-supplemented diet, foci of hepatocyte differentiation emerge by about 12 weeks. The present study deals with pancreatic changes during copper depletion and determines the extent of regenerative capacity of the pancreas during the early phases of copper supplementation. During the period of copper depletion, the pancreas gradually decreased in size, and by eight weeks of deficiency weighed approx. 30% as much as the control pancreas. Light-microscopic examination showed focal necrosis of acinar cells at 4 weeks, loss of lobular architecture by six weeks and loss of 85-90% of acinar tissue by eight weeks. Regeneration of the pancreas was initiated by feeding copper-supplemented diet at the end of eight weeks of copper deficiency. The wet weight of the pancreas increased gradually, and by 17 days it weighed 50% more than on the first day of regeneration. Mitotic activity was observed mainly in the acinar cells, beginning at 24 h and reaching a maximum of 19 +/- 1/1000 acinar cells at 72 h, and decreasing steadily thereafter. [3H]Thymidine autoradiography showed a labeling index of 51 +/- 1.7/1000 acinar cell nuclei at 24 h, reaching a peak value of 261 +/- 5.5/1000 acinar cells at 96 h. By 17 days, pancreatic regeneration was only partially complete; however, the proliferative activity was still persistent, albeit at a slower pace. Morphometric analysis of the pancreas on the 17th day of regeneration showed that acinar tissue occupied only 37% of the volume, as compared to 84% in the normal pancreas. These studies clearly demonstrate that acinar tissue of the pancreas following copper deficiency is capable of regeneration, but that the recovery of the pancreas is only partial. Additional studies are necessary to establish the role of this early acinar cell regeneration in pancreatic hepatocyte differentiation.

Acute toxicity of nimbolide and nimbic acid in mice, rats and hamsters

Nimbolide and nimbic acid are toxic to mice only when given i.p. and i.v. but they are less toxic to rats and hamsters. The LD 50 values of a single i.p. administration of nimbolide to adult male, female and weanling mice were 225, 280 and 240 mg/kg body wt, respectively, and its i.v. LD 50 value was decreased to 24 mg/kg body wt in adult male mice. No fatality was observed when nimbolide was given i.g., i.m. and s.c. to adult male mice. Estimated LD 50 values of nimbolide in rats and hamsters were somewhat higher than 600 and 500 mg/kg body wt. After 12–23 h i.p. administration of a lethal dose, most animals died of possible dysfunctions in kidney (tubular necrosis), small intestine (hemorrhagic necrosis), pancreas (acinar cell necrosis) and liver (mild fatty infiltration and focal necrosis). In contrast, mice and rats given a lethal dose of nimbolide (i.v.) died of a marked and sudden drop in arterial blood pressure and respiratory paralysis within about 1–18 min. Nimbic acid was less toxic to mice with an i.v. LD 50 value of 265 mg/kg body wt and i.p. and i.g. LD 50 values of higher than 600 mg/kg body wt. The possible cause of death induced by nimbic acid may be similar to that of nimbolide given i.v. and this is a sudden hypotensive shock.

Human acute pancreatitis: its pathogenesis in the light of immunocytochemical and ultrastructural findings in acinar cells.

Human acute pancreatitis results from an autodigestive process frequently associated with alcohol abuse, gall stone disease and shock. Peripancreatic fat necrosis was identified as one of the earliest visible lesions, whereas acinar cell necrosis and haemorrhage were regarded as secondary changes. To examine the alterations in acinar cells in more detail, their enzyme content and fine structural features were studied immunocytochemically using antisera against alpha-amylase, lipase, trypsin, chymotrypsin and pancreatic stone protein, and electronmicroscopically in pancreatic tissues from patients with severe acute pancreatitis. Peripheral acinar cells in the immediate vicinity of fat necrosis were found to be heavily degranulated, while acinar cells at some distance of necrosis fully retained their enzyme content. Other frequent changes of the acinar cells included cuboidal transformation, loss of microvilli, increased occurrence of autophagosomes, and formation of enlarged acinar lumina. As there was no apparent cell membrane leakage or rupture of duct lumina, it is concluded that the acinar cells adjacent to fat necrosis release their granules by undirected basolateral extrusion. The findings thus suggest that one of the basic defects in acute pancreatitis is the uncontrolled release of enzymes from peripheral acinar cells into the interstitial space which, in turn, presumably by the action of lipase, leads to autodigestive fat necrosis.

Pancreatic damage produced by injecting excess lysine in rats

Intraperitoneal injection of lysine (400 mg/100 g body weight) in rats caused necrosis of pancreatic acinar cells with fat necrosis and a significant increase in serum amylase and lipase. The early morphological changes in the pancreas were investigated. At 3 to 6 h, marked swelling of mitochondria was observed throughout the cytoplasm followed later by dilation of the endoplasmic reticulum and the formation of autophagic vacuoles, indicative of rapid cellular degeneration. These results suggest that transient disturbance of energy formation following mitochondrial swelling resulted in disorders of protein metabolism, with disorganization of the endoplasmic reticulum and pyknosis of the nuclei as later events.

Caerulein-Induced Acute Necrotizing Pancreatitis in Mice; Protective Effects of Proglumide Benzotript, and Secretin

The onset, course, and regression of the biochemical and structural alterations associated with pancreatitis induced by various doses of caerulein were studied in the mouse. In addition, the protective effect of secretin was compared with that of the cholecystokinin-receptor antagonists proglumide and benzotript. Subcutaneous or intraperitoneal injections of caerulein induced increases in serum amylase concentration and pancreatic weight and histologic evidence of acute pancreatitis, all effects being dose-related. Cytoplasmic vacuoles were the earliest histologic alterations. As the pancreatitis progressed these vacuoles increased to an enormous size. Interstitial inflammation and acinar cell necrosis were prominent after 6 h, reached a maximum after 12 h, and mostly disappeared after 4 days. During the course of pancreatitis approximately 40% of the acinar cells showed signs of severe degeneration or necrosis at the most effective doses of caerulein. Electron microscopy showed both intact and degenerating granules inside the vacuoles. Signs of basolateral exocytosis of zymogen granules were not observed. During the regression of pancreatitis, focal atrophy was a remarkable histologic finding. Repetitive initiation of pancreatitis (six courses of caerulein injections over 5 wk) produced marked focal atrophy and early fibrosis. High doses of proglumide or benzotript markedly ameliorated both the biochemical and structural alterations induced by caerulein. Secretin, even at very high doses, had only minor protective effects. This study presents a model of acute necrotizing pancreatitis in which the severity of the induced pancreatitis ranges dose-dependently from mild interstitial inflammation to severe necrosis. The ultrastructural alterations described herein support the hypothesis that the trigger mechanism of acute pancreatitis appears to be a primary intracellular event rather than an interstitial event that secondarily damages the acinar cells.

What's new in in vitro studies of exocrine pancreatic cell injury?

Exocrine pancreas in vitro models are useful for the study of pancreatic differentiation, secretion mechanisms, cell injury, and lysosomal processing of secretory product. Syrian hamster pancreas in explant organ culture undergoes a series of morphologic changes which parallel in vitro acinar cell injury, differentiation, and phenotypic alteration. Within 48 hours, the cultured acinar cells show morphologic evidence of sublethal cell injury. Autophagy and crinophagy are particularly striking. The autophagic processes can be inhibited by the addition of the protein synthesis inhibitor cycloheximide or by culture at lowered temperatures (20 degrees C). Acinar cells lethally damaged show pyknotic nuclei, high amplitude swelling, and necrosis. Approximately 25% of each explant is viable after 72 hr in culture and the viability remains constant at 25-35% for up to 60 days of culture. The morphological changes of the explants are consistent with many of the features of pancreatitis and carcinoma of the exocrine pancreas. There is an increase in the ductal elements and a decrease in acini over time in culture. This may be due to: (a) an increased replication of ductal epithelial cells concomitant with necrosis of acinar epithelial cells and/or (b) phenotypic alteration of acinar cells to ductal cells. Acinar cell necrosis and phenotypic alterations may in part be due to the activation of lysosomal degradation pathways. Processes which inhibit lysosomal activation proved protective against these alterations, while processes which promote zymogen activation were deleterious.

Pathologic effects of fractionated fast neutrons or photons on the pancreas, pylorus and duodenum of dogs

Thirty-nine adult male Beagles received either fast neutron or photon irradiation to the right thorax to determine the relative biological effectiveness (RBE) of fast neutrons on normal pulmonary tissue. The right anterior abdomen was included in the field of radiation. Twenty-four dogs (six/group) received fast neutrons with an average energy of 15 MeV to total doses of 1000, 1500, 2250 or 3375 rad in four fractions per week for six weeks. Fifteen dogs received 3000, 4500 or 6750 rad of photons (five/group) in an identical fractionation pattern. All neutron irradiated dogs receiving 3375 and 2250 rad and one receiving 1500 rad developed clinical signs of pancreatic, hepatic and gastrointestinal disturbances. The liver enzymes of these dogs became elevated and they died or were euthanatized in extremis 47–367 days after irradiation. Only one 6750 rad photon dog developed similar signs and died 708 days post-irradiation. Five neutron and 10 photon exposed dogs died of other causes. Neutron-induced lesions in the stomach and duodenum included hemorrhages, erosions, ulcerations and fibrosis. Ulcers perforated the GI tract of five dogs. Pancreatic lesions included degranulation and necrosis of acinar cells, fibrosis and atrophy. Islet cells were not obviously damaged. All lesions were associated with degenerative and occlusive vascular changes. The RBE of fast neutrons, assessed by clinical signs, gross and microscopic pathology, is approximately 3–4.5 for pancreas and about 4.5 for pylorus and duodenum.

Noncitrinin Toxicity of Penicillium citrinum Contaminated Corn

Day-old White Leghorn chicks were fed P. citrinum-contaminated corn at 3.88, 7.75, 31, or 62% levels in the ration for a period of 5 weeks. High treatment levels (31 and 62%) of contaminated corn in the ration caused severe growth depression, high mortality, and significant decreases in feed consumption. Necrosis of periportal and centrilobular hepatocytes, glomerular atrophy and hyperplasia, degeneration and necrosis of tubular epithelial cells, lymphoid depletion, suppression of hematopoiesis, necrosis of pancreatic acinar cells, and a cardiac and skeletal myopathy were observed microscopically in chicks fed 31 and 62% contaminated corn in the rations. Low levels of contaminated corn in the ration (3.88 and 7.75%) caused few clinical effects; however, liver and kidney changes were observed microscopically. Ultrastructural observation of affected tissues showed swelling of mitochondria, dilatation of the endoplasmic reticulum, increased numbers of autophagic vacuoles and paramyelin figures, and lipoprotein peroxidation of organelles. The toxin (s) present in the P. citrinum contaminated ration have not been identified.

Experimental pancreatitis in the rat. Light and electron microscopical observations on early pancreatic lesions induced by intraductal injection of trypsin, phospholipase A2, lysolecithin and non-ionic detergent

Trypsin, phospholipase A2, lysolecithin or non-ionic detergent polyoxyethylene p-t-octyl phenol solutions were injected into the rat biliopancreatic duct. Histological and ultrastructural changes in the gland were studied 15 min and 3 h after the injections. The rough surfaced endoplasmic reticulum disintegrated in two ways: (1) the endoplasmic reticulum in the cell periphery was vesiculated but ribosomes were well preserved at 15 min, and (2) large, round membranous structures appeared in apical cytoplasm at 3 h. Zymogen granules disintegrated in the second type, which possibly represents autodigestion. Both types of injury lead ultimately to structureless necrosis. Lesions induced by phospholipase A2 and lysolecithin were identical. Trypsin-induced damage developed slowly and the two phases of endoplasmic reticulum disintegration were not sharply separable. Lesions caused by polyoxyethylene p-t-octyl phenol were variable at 15 min, but at 3 h the type 2 injury described above was observed. It was concluded that although the initial damage in pancreatic acinar cells may vary, necrotic changes are similar despite the injected material at the later time interval. During acute pancreatitis, the acinar cell necrosis is most probably due to the action of lysolecithin produced by the activation of phospholipase A2.

Pancreatic uptake of 99Tc stannous pyrophosphate in experimental viral pancreatitis in mice.

The pancreatic uptake of 99mTc stannous pyrophosphate (99mTc-PYP) in acute pancreatitis in mice induced by coxsackievirus B3 was studied. 99mTc-PYP uptake ratio, measured by the ratio of counts/min/g for the pancreas to counts/min/g for the skull, began to increase 2 days after virus inoculation and markedly increased on day 3 when necrosis of pancreatic acinar cells was evident. On day 5, 99mTc-PYP ratio reached a maximum (2.95 +/- 1.85, mean +/- SD) and histologically fine granules of calcification were seen in the necrotic acinar cells. Thereafter, cellular infiltrations increased and were most severe on day 7, but 99mTc-PYP ratio had began to decrease. On day 28, the necrotic areas of the pancreas were replaced with fatty cells and calcification was still present, though 99mTc-PYP ratio had decreased. The present findings may provide a basis for 99mTc-PYP imaging in cases of clinical viral pancreatitis.

Irradiated salivary glands in the rhesus monkey. A light microscopic study.

Female rhesus monkeys (Macaca mulatta) received fractionated doses of orthovoltage irradiation to the submandibular and sublingual glands. Changes in the glands varied from serous cell degranulation and degeneration in the submandibular glands to acinar cell necrosis and fibrosis in the sublingual glands. Acute inflammation was absent in all irradiated glands. In all glands, the microvasculature appeared normal.

Early membrane injury in lethally irradiated salivary gland cells.

The early manifestations of radiation injury in salivary glands were investigated in the rat. The animals received a single X-ray dose in the range of 200-2000 rad to their neck area. Glandular changes during the first 24 hours were studied by light and electron microscopy and by measuring serum amylase activity. The amount of cell necrosis was quantitated and expressed as necrosis index (NI), Parotid NI and serum amylase activity 24 hours following irradiation were directly proportional to the X-ray dose. The submandibular gland cells were radioresistant and so were the mucous cells of the sublingual gland. The major increase in parotid acinar cell necrosis occurred between 12 and 24 hours after irradiation. However, more than 100 per cent increase in serum amylase level was detected prior to the onset of any significant cell necrosis. As early as two hours following irradiation signs of cell membrane injury were demonstrable in the parotid by electron microscopy and consisted of intracellular oedema, sequestered degenerative cell membranes, and an accumulation of intramitochondrial particles. None of these changes was detectable in the submandibular gland. The implication of membrane injury in the lethal effects of radiation on parotid cells is discussed.