A study of the microcirculation in whole villi of neonatal mice using a peroxidase histochemical staining method.

The anatomy of the microcirculation of intestinal villi from the upper, middle, and lower small intestine of neonatal mice from 8 to 14 days old was studied using a histochemical peroxidase technique that specifically stained erythrocytes. Over 8-14 days, there was little chronological variation between the same regions of gut; the exception was the lower intestine, which, in younger mice, was noticeably less well perfused with erythrocytes. Vascular beds in the middle and lower intestine comprised a hairpin loop with cross-connections. In the upper intestine, the capillary beds were generally more complex, particularly in apical regions of the villi. Most villi were well perfused with erythrocytes, but a minority (less than 10%) contained considerably fewer red cells, some to the point of being totally ischemic. Other villi (less than 5%) were hyperemic, and the vascular beds packed and engorged with red cells. Usually, the packing density (hematocrit) of red cells within blood vessels increased progressively from villus base to apex. Red-cell deformation was more prevalent at villus apices, with marked crenation in some villi, yet in the basal regions of these same villi, the red cells were of normal discoid shape. The peroxidase staining technique produces a reliable histological picture of red cells circulating through villi. It also reveals differential perfusion of erythrocytes between and within villi, and that blood vessels pass through hypertonic zones in the apical regions of villi.

Rotavirus-induced changes in the microcirculation of intestinal villi of neonatal mice in relation to the induction and persistence of diarrhea.

Using a histochemical peroxidase technique, under conditions that preferentially stain erythrocytes, we have shown changes in the microcirculation of villi of neonatal mice infected with murine rotavirus. Between 18 and 48 h postinfection (PI), throughout all areas of the small intestine there occurred, sequentially, a marked ischemia and atrophy of villi. By 72 h PI, villi had recovered their normal height and showed incipient hyperemic microcirculation. At 96 h PI, hyperemic microcirculation was most marked. Between 120 and 144 h PI, a second phase of villus atrophy occurred, which was more attenuated and confined to the upper and middle regions of the intestine. This phase was not accompanied by a wide-spread ischemia of villi: a minority of villi were short and ischemic but many appeared hyperemic. Recovery of villus microcirculation occurred at 168 h PI, which coincided with recovery from diarrhea. These changes in villus microcirculation are discussed in relation to the pathology and pathophysiology of rotavirus infection. We make two novel suggestions. First, the reduction in red cells flowing through villi in the early stages of the infection instigates hypoxia and hence atrophy of villi. The ensuing but ephemeral increase in rate of cell division, necessary for the reconstitution of villi, induces hypersecretion. Second, the increase in numbers of erythrocytes found in villi during their regrowth phase and throughout the remaining time course of the infection perturbs the countercurrent system, lowering the osmolality of the hyperosmotic zone located at villus tips, thereby impairing water absorption and prolonging diarrhea.

X-ray microanalysis of rotavirus-infected mouse intestine: a new concept of diarrhoeal secretion.

Neonatal mice were infected at 7 days of age with rotavirus [epizootic diarrhea of infant mice (EDIM) virus] and killed at 24-h intervals postinfection (PI). Cytoplasmic concentrations of Na, Mg, P, S, Cl, K, and Ca intestinal epithelial cells from infected and age-matched control animals were measured by x-ray microanalysis. In villus tip cells, Ca concentration increased at 24-96 h PI; Na concentration increased at 24-72 h PI; Ca and Na concentrations were near normal by 168 h PI. K concentration decreased 24-72 h PI, and Cl concentration decreased 48-96 h PI. In crypt cells, changes were observed without a discernible pattern: at 96 h PI, Na, Mg, S, and Cl concentrations increased and K concentration decreased; at 120 h PI, the concentrations of all elements except Na and Ca increased. In villus base cells, the mean concentrations of all elements except Ca peaked at 48-72 h PI and at 120 h PI. Na and Cl concentrations increased dramatically in some cells from 48 h PI onward. All the above concentration values were obtained from freeze-dried specimens and expressed in millimoles per kilogram of dry weight. Conversion of a limited number of data, pertaining to villus base cells, from dry weight to wet weight was possible. This conversion revealed that villus base cells in infected animals were more hydrated than corresponding cells from control animals. Also, the Na and Cl concentrations in mmol/kg H2O were significantly higher in villus base cells from infected animals than in those from corresponding controls: 137 +/- 7 versus 38 +/- 4 (Na) and 121 +/- 5 versus 89 +/- 6 (Cl). Wet weight concentrations of other elements were either the same (Mg) or lower (P, S, and K) after infection with virus. From these studies, a new concept of the pathophysiology of rotavirus-induced diarrhoeal secretion is proposed: stimulation of villus base cells to rapid cell division is accompanied by transient accumulation of Na and Cl; excess NaCl is secreted into the lumen, which is the driving force for fluid loss.

Studies on early association of Salmonella typhimurium with intestinal mucosa in vivo and in vitro: relationship to virulence.

The abilities of six strains of Salmonella typhimurium to associate with rabbit ileal mucosa have been measured in vitro. Two were "virulent" strains (TML and W118 which are invasive and inducers of fluid secretion in rabbit ileal loops); four were "avirulent" (LT7, M206 and SL1027 which are invasive but induce negligible fluid secretion, and Thax-1 which is neither invasive nor an inducer of fluid secretion). A special organ-culture apparatus was designed to expose only the luminal surface of the mucosa to organisms. Viable counts of washed homogenised tissue taken 30 min after challenge showed that virulent strains TML and W118 and avirulent strains LT7 and M206 could not be distinguished from each other. Avirulent strain SL1027 associated less well than the other four strains, and Thax-1 associated less well than SL1027; both these strains were non-motile whereas the other four were motile. Thus, early association with gut mucosa did not discriminate all avirulent strains from the virulent strains. Qualitative examination of tissues by scanning electronmicroscopy did not detect strains LT7 and M206 on the mucosal surface whereas strains TML and W118 were readily seen, suggesting that the nature of association of virulent and avirulent strains was different. Qualitative examination by transmission electronmicroscopy of tissues challenged in vivo for 120 min showed virulent strains TML and W118 invading epithelial cells; similar events were reproduced after 120-min challenge in vitro. In contrast, invasion by avirulent strains was observed only very rarely.

An electron microscopic investigation of time-related changes in the intestine of neonatal mice infected with murine rotavirus.

Seven-day-old mice were infected orally with murine rotavirus (EDIM) and regions of the gut examined at 24 h intervals up to 7 days by electron microscopy. Structural changes were correlated with data on viral antigen production, thymidine kinase activity, and clinical signs of diarrhea. No pathological changes were detected in the colon. Infection and structural damage were confined to the small intestine, with middle regions showing the most pronounced changes. Constriction of villus bases, edema of the lamina propria, and vacuolation of enterocytes occurred at 24 h postinfection (PI), i.e., before evidence of major virus replication. Transient villus atrophy occurred at 48 h PI. Recovery of villus length was evident by 72 h PI accompanied by evidence of marked enterocyte replication at villus bases. Many enterocytes were damaged with little evidence for the presence of virus particles. By 96 h PI, villi had almost recovered from infection although some enterocytes were still damaged; no virus particles were detected in these cells. A second phase of villus damage and edema of the lamina propria occurred at 120 h PI; the pathology resembled that at 24-48 h PI. By 144 to 168 h PI, recovery of the mucosa from infection was virtually complete. We suggest that many of the pathological features following rotavirus infection result from rotavirus-induced ischemia of villi and that diarrhea results from malabsorption of fluid by damaged villi and hypersecretion of ions released from increased numbers of dividing cells at villus-crypt borders.

Intestinal enzyme profiles in normal and rotavirus-infected mice.

To investigate further the pathophysiology of rotavirus-induced diarrhea, changes in specific activities of eight relevant intestinal enzymes [alkaline phosphatase, thymidine kinase, lactase, maltase, sucrase, Na+,K+-adenosine triphosphatase (ATPase), adenylate and guanylate cyclases] were measured following infection of suckling mice with murine rotavirus (epizootic diarrhea of infant mouse strain) and compared with age-matched control mice. The concentration of lactose within the lumen of the gastrointestinal tract during infection was also measured. During the course of infection, activities of alkaline phosphatase and lactase decreased, whilst the activity of thymidine kinase increased. Precocious maturation profiles of sucrase and maltase enzymes were observed. No significant changes were detected in the activities of Na+,K+-ATPase or the adenylate and guanylate cyclases. These results are discussed in relation to existing and novel hypotheses on the pathogenesis of rotavirus-induced diarrhea.

Expression of an antigen in strains of Salmonella typhimurium which reacts with antibodies to cholera toxin.

Six strains of Salmonella typhimurium (W118, TML, SL1027, LT7, M206 and Thax 1) of different virulence were examined for the presence of antigens which react with antibodies to cholera toxin (anti-CT). A fluorescent-antibody-labelling technique employing anti-CT was used to analyse antigen expression. A rapid increase in the proportion of cells producing a CT-related antigen was demonstrated in cells in early log phase (1-4 h growth) followed by a rapid decline during mid-late log phase in each of the six strains. The nature of the CT-related antigen was analysed by immunoblotting using anti-CT. An antigen of mol. wt equivalent to a high-mol. wt species of CT B subunit was detected in polymyxin-B extracts of all strains but greater amounts were observed in the strains that we consider avirulent. Nothing equivalent to a CT A-related subunit was observed in any of the strains. The relatedness of the salmonella antigen to CT was limited. The high-mol. wt antigen was not disrupted in the denaturing conditions of SDS-PAGE; nothing was detected by enzyme-linked immunosorbent assays with either ganglioside or anti-CT as anchor.

Kinetics, tissue specificity and pathological changes in murine rotavirus infection of mice.

Mice that did not contain antibodies to rotavirus were orally infected with murine rotavirus (EDIM strain) and observed over 7 days. As judged by ELISA, only the small intestine was infected, not the colon. The infection was biphasic, viral antigen peaks being observed at 48 h and approximately 120 h post-infection. Clinically evident diarrhoea was maximal at 72 h. Virus in the upper, middle and lower regions of the small intestine was mainly tissue-associated; most virus was found in the middle small intestine. Two peaks (48 h and 120 h post-infection) of virus antigen were observed in the colon, but these corresponded to luminal, not tissue-associated viral antigen. Only enterocytes in the upper two-thirds of villus epithelia were infected as judged by fluorescent-antibody analysis and transmission electron microscopy. Scanning electron microscopy revealed morphological appearances not hitherto correlated with the progress of the infection: villus tips were convoluted, corresponding to the shedding of virus-infected cells but the lower regions of infected villi were shrunken and considerably narrowed compared to tips.

The nature and role of mucosal damage in relation to Salmonella typhimurium-induced fluid secretion in the rabbit ileum.

The time course and nature of mucosal damage induced in rabbit ileal loops by two strains of Salmonella typhimurium (TML and W118) isolated from human infections was assessed by immunofluorescence microscopy and by scanning and transmission electronmicroscopy. Salmonella-induced fluid secretion occurred in the presence or absence of gross mucosal architectural damage. Neither strain caused mucosal ulceration. When damage did occur, the villi were shortened by loss of their tip regions with concomitant reforming of an intact mucosal surface. Immediately preceding the onset of fluid secretion, marked infiltration of the mucosa with polymorphonuclear leukocytes and occasional macrophages was seen. This revives an earlier suggestion that interaction between invading salmonellae and acute inflammatory cells may be an important factor in initiation of fluid secretion. Brush-border invasion by salmonellae cannot per se be the immediate cause of fluid secretion, because the latter occurred several hours after initial invasion.

Enterotoxin production by Salmonella typhimurium strains of different virulence.

Six strains of Salmonella typhimurium (TML, W118, LT7, SL1027, M206 and Thax-1) of known virulence and ability to induce fluid secretion when inoculated into the rabbit ileum were examined for enterotoxin production. Enterotoxic activity, assayed in the rabbit ileal-loop test, was detected in polymyxin-B extracts from all strains (with the possible exception of Thax-1) cultured for 6 h in casamino acid-yeast extract medium. The extracts were inactive in tissue-culture assays with CHO, Y-1 adrenal and Vero cells, and in the infant mouse assay for enterotoxin. There was no correlation between enterotoxigenicity in vitro and the ability of whole organisms to induce fluid secretion in vivo. The significance of these results in relation to salmonellosis is discussed.