Haemopoietic tissues exposed to ionizing radiation are shown to exhibit increased macrophage activation, de®ned by ultrastructural characteristics and increased lysosomal and nitric oxide synthase enzyme activities. Macrophage activation post-irradiation was also associated with enhanced respiratory burst activities and an unexpected neutrophil in®ltration. Examination of p53-null mice demonstrated that macrophage activation and neutrophil in®ltration were not direct e ects of irradiation, but were a consequence of the recognition and clearance of radiation-induced apoptotic cells. Increased phagocytic cell activity was maintained after apoptotic bodies had been removed. These ®ndings demonstrate that, contrary to expectation, recognition and clearance of apoptotic cells after exposure to radiation produces both a persistent macrophage activation and an in¯am-matory-type response. We also demonstrate a complexity of macrophage activation following radiation that is genotype dependent, indicating that the in vivo macrophage responses to radiation damage are genetically modi®ed processes. These short-term responses of macrophages to radiation-induced apoptosis and their genetic modi®cation are likely to be important determinants of the longer-term consequences of radiation exposure. Furthermore, in addition to any e ects attributable to immediate radiation-induced damage, our ®ndings provide a mechanism for the production of damage via a`bystander' e ect which may contribute to radiation-induced genomic instability and leukaemogenesis. Oncogene (2001) 20, 7085 ± 7095.

Human tissues, both biopsy and postmortem, and tissues from rodents were fixed by microwaves at various temperatures and compared against formaldehyde-fixed material. Conventional stains, including trichromes, worked well. Red cells were lysed, but white cells were fixed, thus permitting diagnoses of various inflammatory states. Malignant cells were equally well-preserved by the two methods. Histochemical investigations of mucosubstances, lipids and various hydrolases showed no significant difference between the two techniques. Some neurological stains, however, were not as good following microwave treatment. Immunocytochemical localization of IgA, IgM and IgG showed no significant difference after microwave fixation compared to that in tissues fixed with formaldehyde. Microwave fixation did not lead to a greater tissue shrinkage than that obtained with formaldehyde fixation. Both were significantly less than that following treatment with phosphate-buffered saline alone. Electron microscopy gave results which were interpretable, but with damage resembling early postmortem change. Microwave fixation is complete in approximately 1-2 min. The mechanism of fixation appears to be due to denaturation associated with disulphide bond formation and a decrease in solubility of proteins.

SUMMARY Unequivocal apoptoses were seen by light microscopy in examples of leprosy, sarcoidosis, tuberculosis, Crohn's disease and foreign body granulomata. A limited electron microscopic investigation showed typical apoptotic bodies in both sarcoid and leprosy granulomata. The number of apoptoses and mitoses in granulomata were counted and their densities calculated. The wide variation in the results between individual lesions may reflect differences in disease activity.Current understanding of cell turnover in granulomata is based largely on studies of experimentally induced granulomata in animals by Dannenberg, Spector, and others.' These experiments established that mononuclear phagocytes in granulomata have a finite life span and that this varies with the eliciting agent. The eventual fate of mononuclear cells in granulomata is not clear, although evidence from some studies suggests that cell death has a major role.2 Cell necrosis is not readily apparent in noncaseating granulomata, however, and hence regression of differentiated mononuclear phagocytes to less mature forms with subsequent emigration from the lesion has been proposed as an alternative explanation.3Another possible mechanism for the loss of mononuclear phagocytes from granulomata is apoptosis, a morphologically distinctive form of cell death which affects isolated cells in living tissue.4 Apoptosis occurs in a wide range of physiological and pathological circumstances, including involution of the adrenal cortex,5 secretory endometrium,6 tadpole tails,7 and basal cell carcinoma.8 Cells undergoing apoptosis show condensation and fragmentation into a number of membrane bound eosinophilic fragments which are then endocytosed by surrounding cells, even if these are not normally phagocytic.9 In contrast with true necrosis, cells undergoing apoptosis do not provoke an inflammatory response. As the process lasts a matter of hours, detection of even small numbers of apoptotic bodies within tissues indicates appreciable cell loss.During a recent study of the histology of leprosy lesions, we observed apoptoses within both tuberculoid and lepromatous lesions.10 We now have ex-

SUMMARY Oesophageal mucosal biopsies were incubated in 20, 2, and 0.2 mM solutions of cholic, chenodeoxycholic, ursodeoxycholic, and deoxycholic bile acids. Both conjugated and unconjugated bile acids were studied at pH 1 and 7 singly and in combination. Observations were also made using 0.1 N hydrochloric acid and human gastric juice at pH 1-3 and 7-8. After incubation for up to 15 minutes the mucosa was examined under transmission electron microscopy. We concluded that high and moderate concentrations of all the common bile acids damaged the oesophagus irrespective of the pH, that low concentrations of bile acids were damaging only at high acid levels, and that damage to the epithelium did not occur when the pH of the gastric juice had been raised.Oesophagitis is a common problem which is related to retrograde reflux of gastrointestinal secretions.' 4 We have developed a system of incubation of mucosal biopsies to examine which components of gastrointestinal fluids are important in causing damage,56 and have demonstrated that many constituents could be toxic.5 The roles of bile acids and hydrogen ions are particularly important because they are major components of refluxing material. H2-receptor antagonists effectively reduce the hydrogen ion content of gastric secretion and thus the composition of material to which the oesophagus is imposed during an episode of reflux. It was therefore decided to study the effects of different physiological concentrations of human bile acids and the effects of both high and low hydrogen ion concentration. Methods PATIENTSThirty-five patients were studied. All were undergoing investigation by upper gastrointestinal endoscopy for abdominal symptoms. The patients were taking a variety of drugs all of which were omitted for six hours before the endoscopy. Four patients were receiving a course of therapy with cimetidine. None of the patients had endoscopic evidence of oesophagitis. After the endoscopic examination had been completed using a forward-viewing instrument, Received for publication 18 November 1980 the endoscope was withdrawn to a point 10 cm proximal to the apparent junction of the oesophageal and gastric mucosa and multiple biopsies were taken with the endoscopy forceps. The biopsies were incubated immediately at 37°C in the test solutions for five, 10, or 15 minutes, and then immersed in 3 % glutaraldehyde at 4°C at pH 7.2 in 0 1 M cacodylate. To confirm the normality of the oesophageal mucosa, control biopsies taken at the same time were immediately fixed in glutaraldehyde. All the samples were post-fixed with osmium tetroxide, dehydrated, and embedded in Araldite. Thin sections were cut, stained with uranyl acetate and lead nitrate, and examined with a Jeol Cx100 electron microscope at 60 KV.All the incubations were performed in duplicate, triplicate, or quadruplicate. The incubation mixtures were as follows, with number of patients studied in parentheses.1. Model conjugated bile acid solutions at pH 7, in concentrations of 20, 2, and 0-2 mM, containing a mixtur...