Gnotobiotic Piglets Experimentally Infected with Neonatal Calf Diarrhoea Reovirus-Like Agent (Rotavirus)

Transcription

1 Vet. Pathol. 13: (1976) Gnotobiotic Piglets Experimentally Infected with Neonatal Calf Diarrhoea Reovirus-Like Agent (Rotavirus) G. A. HALL, JANICE C. BRIDGER, R. L. CHANDLER and G. N. WWDE ARC Institute for Research on Animal Diseases, Compton, Newbury, Berkshire Abstract. Gnotobiotic piglets were infected with a strain of rotavirus that had been isolated from a calf and passaged four times in piglets. The resulting disease was studied by light and electron microscopy at 21,44 and 68h after inoculation. In thesmall intestine infection caused desquamation of the epithelial cells of the villi resulting in severe stunting. There was severe damage to microvilli and accumulation of lipid within the cytoplasm. Virus particles were seen in epithelial cells covering the stunted villi. Infection also caused desquamation of the superficial epithelial cells of the stomach mucosa and of the epithelial cells of the bronchi, bronchioles, and alveoli of the lungs. The importance of the neonatal calf diarrhoea reovirus-like agent as a cause of diarrhoea in calves was first established in the United States [I 1,241. Subsequently a similar virus was isolated from diarrhoeic calves in Great Britain [25], Australia [22], Canada [15], and Denmark [13]. The reovirus-like particles, which have been seen in faeces of young children with acute nonbacterial gastroenteritis, have been found to be morphologically indistinguishable from those of neonatal calf diarrhoea [8, 101. The calf and human viruses also appeared morphologically and antigenically distinct from both reoviruses and orbiviruses, and the term rotavirus has been proposed [7-91. Although the calf and human rotaviruses are morphologically indistinguishable, we were unable to infect calves with human virus. Piglets, however, are susceptible to infection with human virus [5] but do not become ill. We passed calf rotavirus serially five times in gnotobiotic piglets, and each passage produced profuse diarrhoea, depression and anorexia. Our serological studies have shown that pigs are naturally infected either with calf

2 198 HALL et a/. rotavirus or a related pig virus with similar antigenic properties. We have isolated a virus with similar morphology to the calf rotavirus from weaned pigs with spontaneous diarrhoea. Studies of the pathology of animals infected with rotavirus have been confined to calves [12, 15,181 and mice [ I, 21. This paper describes the pathological changes in gnotobiotic piglets experimentally infected with a strain of rotavirus originally isolated from a calf. Materials and Methods A litter of eight 6-day-old gnotobiotic piglets was produced and raised as described [20]. Six piglets were infected and two each were killed at 21,44 and 68 h after inoculation. One uninoculated control piglet was killed at 21 h and another at 68 h after infection. A filtrate of intestinal contents from a gnotobiotic calf inoculated with a suspension of calf rotavirus [25] constituted the initial inoculum for four serial passages in gnotobiotic piglets. The six piglets were inoculated intranasally with 2.0 ml of a bacteria-free filtrate from the fourth pig passage. Under pentobarbitone sodium anaesthesia paired segments of upper, middle and lower small intestines were tied off and sufficient fixative inoculated into the tied-off segments to cause slight distension. One segment of each pair was inoculated with 12% neutral buffered formalin and the other with 3% phosphate-buffered glutaraldehyde [17]. When the last segment had been inoculated the piglet was killed without delay using pentobarbitone sodium, and the tied off segments were immediately removed and immersed in fixative. The carcass was then examined. Samples of lung, liver, kidney, urinary bladder, stomach, caecum, and colon were fixed by immersion in 12% neutral buffered formalin. The tied off segments were immersed in fixative for 1-2 h, and then blocks of suitable size were cut and immersed in fresh fixative. On the following day glutaraldehyde-fixed tissue was immersed for 2 h in I.O% phosphate-buffered osmium tetroxide solution, ph 7.3 [14], dehydrated through graded alcohols and the blocks embedded in araldite with the villi oriented for longitudinal sectioning. From these blocks I.O-pm sections were cut, stained with toluidine blue and examined with a light microscope to check the orientation. Ultra-thin sections were cut with a Reichert OMU2 ultramicrotome, stained with uranyl acetate and lead citrate, and examined using a Philips EM 300 electron microscope with an accelerating voltage of 80 kv. Blocks of formalin-fixed material were embedded in paraffin wax, sections cut at 5 pm, and stained with haematoxylin and eosin. Blocks were also embedded in gelatin, and frozen sections were cut at IS pn~ and stained with oil red 0 for neutral fat. Contents of the small intestine were collected at postmortem examination from the unfixed parts of the intestine of each piglet and were centrifuged. After negativestaining they were examined with the electron microscope for virus particles Primary pig kidney and calf kidney cell cultures were inoculated with 1 :3 v/v dilutions of intestinal contents and examined at intervals for the presence of virus by immunofluorescence [4].

3 Rotavirus infection of piglets 199 Results Throughout the experiment the two uninoculated control piglets remained healthy and ate normally. Within 16 h of inoculation the infected piglets had developed diarrhoea and had visibly lost condition. There was a marked reduction in food intake. At 44 and 68 h after inoculation the diarrhoea was less severe, but the piglets were still eating less than normal and were in poor condition. Postmortem examination of the control piglets did not reveal any abnormalities. There was food in the stomachs of both animals, and the contents of the large intestines were solid. The small intestines were pink and the contents semi-solid. Long villi were visible on the mucosal surface. Postmortem examination of the infected piglets showed small amounts of food in the stomach and fluid in the large intestines. In the most severely affected piglets the walls of the small intestine were thin and adhered to the fingers when handled; the lumen was collapsed. In the two piglets killed at 68 h after inoculation dark red areas were seen in the apical and anterior diaphragmatic lobes of both lungs. Lesions were not detected by light microscopy in sections of lung and stomach from control piglets. The villi of the upper small intestine were covered by columnar epithelial cells each of which had a basal nucleus, dense eosinophilic cytoplasm and a well-defined brush border. Epithelial cells covering the upper half of the villi in the middle small intestine contained many intracytoplasmic vacuoles, and the position of the nuclei was apical or medial. Villi from the lower small intestine were also covered by vacuolated cells except at the base of the villi where unvacuolated cells were seen. Neutral fat was absent from epithelial cells of villi in the upper and lower small intestine. In the middle small intestine fat was seen on the surface of the epithelial cells and laterally at the base of the cells. Fat was also in the lamina propria. There were no abnormalities in sections of liver, kidney, urinary bladder, caecum and colon of infected piglets at any stage after infection. In the small intestine infection caused rapid desquamation of the epithelial cells from the upper three-quarters of the villi. Epithelial cells appeared to be desquamated in the absence of the usual signs of necrosis so that epithelial cells with eosinophilic cytoplasm and pyknotic nuclei were not seen on the surface of villi (fig. I, 2). The result of epithelial cell loss was severe stunting of the villi (fig. 3). The epithelial cells left covering the stunted villi were flattened or cuboidal, lacked a brush border and contained numerous vacuoles in the apical cytoplasm (fig. 3), many of which

4

5 Rotavirus infection of piglets 20 I contained neutral fat (fig. 4). Occasionally, the lamina propria was exposed (fig. 2) and contained pyknotic nuclei. The distribution of the changes and an estimate of their severity at different stages after infection are shown in table I. The middle small intestine was most severely affected; there were lesions at 21 h after infection, and stunting of villi was most pronounced at 68 h. The lower small intestine was affected at 21 h but had recovered by 68 h, whereas in the upper small intestine lesions were not seen until 44 h and were still present at 68 h. In the stomach at 21 h there was necrosis and desquamation of the goblet cells of the intercrypt mucosa (fig. 5) without exposure of the lamina propria. Not all the mucosa was affected but it was not possible to detect whether there was any difference in susceptibility between the mucosal cells of the cardiac, fundic and pyloric regions. At 44 and 68 h the epithelial cells of the intercrypt mucosa were deeply staining cuboidal cells that did not contain mucin. Occasionally, areas of mucosa made up of normal mucus-containing cells were seen. In the lungs at 21 h slight thickening of the alveolar walls was the only change, but at 44 and 68 h there were changes in the epithelial cells of the terminal bronchi, bronchioles and alveoli. Many of these cells were degenerate and had desquamated into the lumen. The remaining attached cells were dark-staining and irregularly arranged. In the affected areas there was obliteration of alveolar air spaces due to collapse, after necrosis and desquamation of cells from the alveolar walls, hyperplasia of cells in the alveolar walls, congestion of the alveolar walls and a slight mononuclear cell infiltration of the interstitial tissue. The ultrastructural morphology of epithelial cells from villi of the three areas of small intestine of control pigs was similar to that in previous descriptions [6, 19, 211. The apical surface of these cells was covered by long regularly arranged microvilli. Pinocytotic vesicles and an apical tubular system were in the cytoplasm below the microvilli (fig. 6). In the middle and Fig. I. Villi from middle small intestine of an uninoculated piglet killed 21 h after start of experiment. HE. Fig. 2. Villi from middle small intestine of piglet killed 21 h after inoculation. Epithelial cells desquamated. Lamina propria exposed (arrow). Villi are moderately stunted. HE. Fig. 3. Villus from middle small intestine of piglet killed 44 h after inoculation. Villus is severely stunted, and the covering epithelial cells are cuboidal and vacuolated. HE. Fig. 4. Villus from the middle small intestine of piglet killed 44 h after inoculation. The vacuoles in epithelial cells contain neutral fat. Oil red 0.

6 202 HALL et at. Table I. Changes seen with light and electron microscopes in intestines of piglets experimentaliy infected with rotavirus Killed, Pig Site Light microscopy Electron microscopy hours epithelial stunted pyknotic neutral damag- lipid viral after cell des- villi nuclei in fat in ed parinoculation quama- lamina apical micro- ticles tion propria cytoplasm villi upper middle lower upper middle lower upper middle lower upper middle lower upper middle lower upper middle lower Severity of changes has been estimated: - T - - = very severe, -+ F = severe, i T = moderate, t = slight. * Severity of changes has not been estimated: - = present, - = absent. lower small intestine large vacuoles containing small amounts of electrondense material that did not have the appearance of lipid were seen in the cytoplasm. In the upper and middle segments there were dilations of the intercellular spaces towards the base of the cells, and some of these spaces contained material that looked like lipid. The ultrastructural morphology of many epithelial cells examined from infected piglets was abnormal. Two changes predominated : damage to the microvilli, which were short, misshapen and irregular in spacing and orientation (fig. 7), and lipid material in the apical cytoplasm (fig. 7). In many cells with damaged microvilli the apical tubular system was absent (fig. 7). In infected cells at 68 h there were vacuoles within the cristae of mitochondria (fig. 7).

7 Rotavirus infection of piglets 203 Fig. 5. Stomach mucosa of piglet killed 21 h after inoculation. Superficial epithelial cells are desquamating. HE. Groups of viral particles were seen within defined areas of low electron density in the apical cytoplasm, and these areas resembled vacuoles. Each vacuole had a poorly defined perimeter but was often partly circumscribed by membrane with a related ribosome array (fig. 7, 8). Some groups of viral particles, however, occurred free within the cytoplasm. In some instances, there was a possible association with microtubules, and in one case the staining affinity of the tubules was enhanced. Not all cells containing viral particles had damaged microvilli, and viral particles were not detected in many cells that had damaged microvilli. Particles considered to be mature virions were densely stained and were often associated with less dense particles. The mature viral particles had poorly defined circular profiles with an area of reduced staining towards the centre and a dense central spot (fig. 8). There was no evidence of a welldefined enveloping membrane, and the diameter of the particles, using the outer dense component as a reference, averaged 62.5 nm. In some cells the viral particles evidenced a lattice array, and in these arrays the outer components of the virions were not contiguous. The centre to centre spacing of virions within a lattice array was 83 nm. Masses of electron-dense, amorphous material, not enclosed by membrane, were sometimes seen adjacent to virions (fig. 9). In one cell linear profiles were seen within an area of electron-dense material, and densely staining particles also had a linear arrangement (fig. 10).

8 204 HALL er at. Fig. 6 and 7

9 Rotavirus infection of piglets 205 Fig. 8. Higher magnification of part of figure 7. Viral particles are in a vesiculated area limited not by an intact membrane but by a linear arrangement of ribosomes. The viral particles have a very dense punctate centre, surrounded by small, pale area and outer dense shell; there is no clearly defined outer membrane. The distribution of the ultrastructural changes at different sites and stages after infection is shown in table I. The majority of changed cells were seen at 44 and 68 h after inoculation, and the middle and upper small intestine were most often affected. Virus particles were most easily detected in areas where villi were severely stunted. Fig. 6. Middle small intestine of uninoculated piglet. Microvilli are long and regularly arranged; apical tubular system (T) is below the border. There are junctional complexes towards the apex (A-B). Apical cytoplasm rich in endoplasmic reticulum. Vesicles (V), whose contents are consistent with normal absorption, are common. Mitochondria (M) are normal. Fig. 7. Middle small intestine of piglet killed 68 h after inoculation. Cells are pale, and cytoplasmic matrix is amorphous. Mitochondria are densely stained and vacuolated. Microvilli are reduced in number, short and irregularly arranged, the apical tubular system is absent. Irregularly shaped vesicles (V) are in apical cytoplasm, and some contain densely stained viral particles (P). Large inclusions of medium density (L), resembling lipid, are common.

10 206 HALL er a/. Fig. 9 and 10

11 Rotavirus infection of piglets 207 Virus was not detected in the intestinal contents of the two control pigs either by electron microscopy or cell culture, but was detected by both methods in the intestinal contents of all infected pigs. Discussion Histological and ultrastructural studies have been made of calves [12, 181, mice [I, 21, and infants [3] with acute gastroenteritis associated with rotavirus infection. Lesions seen with the light microscope in the small intestines of calves were similar to lesions in our study. The stunting of villi in our piglets was more severe than that reported in other studies on calves [12,25]. Villi of piglets infected with a rotavirus strain derived from a field case of enteritis in pigs were as severely stunted as the villi in our study in which a calf virus was used [unpublished observations]. The histology of infants infected with rotavirus has not been investigated in detail but stunted villi have been described [3]. In mice, however, stunting of the villi was not seen [1, 161, but epithelial cells at the tips of villi sloughed, and those remaining were vacuolated and had eosinophilic cytoplasmic inclusions. Cytoplasmic inclusions were not reported in calves [12] and were not seen in our piglets. These differences in the severity of the lesions between species may be related to the host, to the infective dose, to the strain of virus, or to the stage of the disease at which samples were taken. The most severe stunting in our piglets was at 68 h after inoculation, 50 h after the onset of diarrhoea, whereas the calves in another study were killed between 19 and 31 h after inoculation, less than 6 h after the onset of diarrhoea [12, 181. The mice were examined 24 h after the onset of diarrhoea. Ultrastructural changes in the epithelial cells covering villi in infected calves were rare [18] but were similar to the changes seen in this study. Intracytoplasmic lipid, however, was not seen in the calves nor was there vacuolation of mitochondria. The ultrastructural changes in infected mice [I] correspond closely with those in our piglets. The occurrence of intracyto- Fig. 9. Apical cytoplasm of a cell with densely stained mature virus particles (P), less densely stained particles (D), a large dense amorphous mass and tubular structures (T). There are also lipid (L) and altered mitochondria (M). Fig. 10. Inclusion (A-B) within the cytoplasm of an epithelial cell from a virus affected area. Structure is crystalline with linear arrays of tubule-like components along which are dense particles with a diameter of 33 nm. A cytosome (C) is also shown.

12 208 HALL et al. plasmic lipid has been described in piglets infected with transmissible gastroenteritis virus [21] and has been interpreted as a breakdown in intracellular transport in virus-infected cells. A similar interpretation is possible for our piglets, and the absence of lipid in infected calf cells may be due to differences in transport mechanisms. In addition to differences in the severity of lesions in calves [12, 181 and piglets, the pathogenic processes were dissimilar. After ingestion of the virus a wave of infection appeared to pass along the intestine of calves. Tall columnar cells became infected with virus, and as they desquamated they were replaced by infection-resistant and virus-free cuboidal cells. In our pigs, however, virus particles were not seen in tall columnar cells early in the infection but only in cuboidal or flattened cells covering the stumps of the villi at 44 and 68 h after inoculation. We cannot say whether these infected cells were surviving cells from the villi that did not desquamate or crypt cells that had become infected. Lesions in stomach and lungs have not been described in rotavirus infections. The stomach lesions were mild but might account for the vomiting we have seen in piglets in other studies with a pig rotavirus. The distribution of the lung lesions suggested direct infection of epithelial cells by virus that had passed down the trachea at the time of inoculation. The lesions were not severe and may have little clinical significance. The morphology of the virus particles is similar between species but there are differences in particle size. In mice [I] most particles measured 86 nm in diameter and had a central core and also an outer membrane that appeared to come from the rough endoplasmic reticulum. A small number of a second type of particle was seen; these particles had an average diameter of 65 nm and lacked the outer membrane. In humans two types of particles were also seen - one with a dense core and less dense outer zone measuring 67 nm in diameter and larger particles with an outer membrane measuring 87 nm in diameter [3]. In calves and pigs only the smaller particles without membrane were seen. These differences in size between viral particles may depend on the process of virus maturation. The larger particles in mice and humans seem to get their outer membrane by passing through the rough endoplasmic reticulum. In calves and pigs smaller particles were in the dilated rough endoplasmic reticulum but had not acquired an outer membrane. This study shows that one strain of rotavirus can cause disease in more than one species. There is evidence that this strain of rotavirus exists naturally as a pathogen in pig herds. Sixty percent of sera from sows in the Institute s herd neutralised the calf rotavirus, and piglets that had suckled colostrum

13 Rotavirus infection of piglets 209 from one of these sows were resistant to infection with the virus [unpublished observations]. Acknowledgements The authors thank Mr. D. LUTHER for cell cultures, Mr. M. DENNIS for care of gnotobiotic animals, Mr. P. DENNIS and Mr. TURFREY for histological sections, Mr. H. ANGER for electron microscopical preparations, and Mr. G. FLETCHER for the photographs. References 1 ADAMS, W. R. and KRAFT, L. M.: Electron-microscopic study of the intestinal epithelium of mice infected with the agent of epizootic diarrhea of infant mice (EDIM virus). Am. J. Path. 51: (1967). 2 BANFIELD, W.G.; KASNIC, G., and BLACKWELL, J.H.: Further observations on the virus of epizootic diarrhea of infant mice. Virology 36: (1968). 3 BISHOP, R. F.; DAVIDSON, G. P.; HOLMES, I.H., and RUCK, B. J.: Virus particles in epithelial cells of duodenal mucosa from children with acute non-bacterial gastroenteritis. Lancet ii: (1973). 4 BRIDGER, J.C. and WOODE, G.N. : Neonatal calf diarrhoea: identification of a reoviruslike agent (rotavirus) in faeces by immunofluorescence and immune electron microscopy. Br. vet. J. 131: (1975). 5 BRIDGER, J.C.; WOODE, G.N.; JONES, J.M.; FLEWETT, T.H.; BRYDEN, A.S., and DAVIES, H.: Transmission of human rotaviruses to gnotobiotic piglets. J. med. Microbiol. 8: (1975). 6 DREW, D.T. and WAXLET, G.L.: Enteric colibacillosis in gnotobiotic swine: an electron microscopic study. Am. J. vet. Res. 31: (1970). 7 FERNELIUS, A.L.; RITCHIE, A. E.; CLASSICK, L. G.; NORMAN, J. O., and MEBUS, C. A.: Cell culture adaptation and propagation of a Reovirus-like agent of calf diarrhea from a field outbreak in Nebraska. Arch. ges Virusforsch. 37: (1972). 8 FLEWETT, T.H.; BRYDEN, A.S.; DAVIES, H.; WOODE, G.N.; BRIDGER, J.C., and DERRICK, J. M.: Relation between viruses from acute gastroenteritis of children and newborn calves. Lancet i: (1974). 9 KAPIKIAN, A.Z.; CLINE, W. L.; MEBUS, C.A.; WYATT, R. G.; KALICA, A. R.; JAMES, H.D.; VANKIRK, D.; CHANOCK, R. M., and KIM, H. W.: New complement-fixation test for the human Reovirus-like agent of infantile gastroenteritis. Lancet i: (1975). 10 KAPIKIAN, A.Z.; KIM, H. W.; WYATT, W.G.; RODRIGUEZ, W. J.; Ross, S.; CLINE, W.L.; PARROTT, R.H., and CHANOCK, R. M.: Reoviruslike agent in stools: association with infantile diarrhoea and development of serologic tests. Science 85: (1974). 1 1 MEBUS, C. A.; UNDERDAHL, N. R. ; RHODES, M. B., and TWIEHAUS, M. J. : Calf diarrhea (scours) reproduced with a virus from a field outbreak. Neb. Agr. exp. Sta., Univ. Neb. Res. Bull. 233 (1969).

14 210 HALL el a/. 12 MEBUS, C.A.; STAIR, E.L.; UNDERDAHL, N. R., and TWIEHAUS, M. J.: Pathology of neonatal calf diarrhoea induced by a reo-like virus. Vet. Pathol. 8: (1971). 13 MEYLING, A.: Reo-like neonatal calf diarrhoea (NCD) virus demonstrated in Denmark. Acta. vet. scand. IS: (1974). 14 MILLONIG, G.: Advantage of a phosphate buffer for osmium tetroxide solutions in fixation. J. appl. Physiol. 32: 1637 (1961). 15 MORIN, M.; LAMOTHE, P.; GAGNON, A,, and MALO, R.: A case of viral neonatal calf diarrhea in a Quebec dairy herd. Can. J. comp. Med. 38: (1974). 16 PAPPENHEIMER, A.M. and ENDERS, J.F.: An epidemic diarrheal disease of suckling mice. 11. Inclusions in the intestinal epithelial cells. J. exp. Med. 85: (1947). 17 SABATINI, D. D.; BENSCH, K., and BARNETT, R. J.: Cytochemistry and electron microscopy. The preservation of cellular ultrastructure and enzymatic activity by aldehyde fixation. J. Cell Biol. 17: (1963). 18 STAIR, E. L.; MEBUS, c. A. ; TWIEHAUS, M. J., and UNDERDAHL, N. R.: Neonatal calf diarrhea: electron microscopy of intestines infected with a reovirus-like agent. Vet. Pathol. 10: (1973). 19 STALEY, T.E.; JONES, E.W., and CORLEY, L.D.: Attachment and penetration of Escherichia coli into intestinal epithelium of the ileum of newborn pigs. Am. J. Path. 56: (1969). 20 TAVERNOR, W.D.; TREXLER, P.C.; VAUGHAN, L.C.; Cox, J.E., and JONES, D.G.C.: The production of gnotobiotic piglets and calves by hysterotomy under general anaesthesia. Vet. Rec. 88: (1971). 21 THAKE, D.C.: Jejunal epithelium in transmissible gastroenteritis of swine. Am. J. Path. 53: (1968). 22 TURNER, A. J.; CAPLE, I.W., and CRAVEN, J.A.: Demonstration of virus particles in intestinal contents of calves with diarrhoea. Aust. vet. J. 49: 544 (1973). 23 WELCH, A. B. and THOMPSON, T. L. : Physicochemical characterization of a neonatal calf diarrhea virus. Can. J. comp. Med. 37/3: (1973). 24 WHITE, R.G.; MEBUS, C.A., and TWIEHAUS, M.J.: Incidence of herds infected with a neonatal calf diarrhea virus. Vet. Med. 65: (1970). 25 WOODE, G.N.; BRIDGER, J.C.; HALL, G.A., and DENNIS, M. J.: The isolation of a reovirus-like agent associated with diarrhoea in colostrum-deprived calves in Great Britain. Res. vet. Sci. 16: (1974). Dr. G.A. HALL, Institute for Research on Animal Diseases, Compton, Newbrrry, Berkshire, RG16 ONN (England)