|
Pathology of calves with diarrhoea in southern Britain
Abstract
Twenty-one moribund calves with diarrhoea were purchased from 11 farms, their faeces examined for enteropathogens and samples of intestinal tissue removed under anaesthesia. Lesions and presence of enteropathogens on the" mucosal surface were scored by histological examination of immunostained paraffin sections. Two or more enteropathogens were detected in 19 calves. Cryptosporidium appeared to be the principal cause of diarrhoea in six calves, rotavirus in four, Salmonella typhimurium in two, bacteria adherent to the surface of the large intestine in two, coronavirus in one and K99+ Escherichiacoli in one calf. Diarrhoea in four calves was the consequence of mixed infections in which no one enteropathogen appeared to predominate. In one calf no enteropathogen was detected. Diarrhoea was associated with infections and lesions throughout the small and large intestines. The enteropathogens most frequently. associated with lesions in the small intestines were rotavirus, coronavirus and cryptosporidium; in the large intestines they were coronavirus and bacteria apparently adherent to the mucosal surface.
THE aetiology of neonatal calf diarrhoea is complex; many infectious agents, singly or in combination, have been associated with field outbreaks (Acres et al 1975 , Morin et al 1976 and non-infectious factors contribute to the disease process (Roy 1980) . The infectious agents thought to contribute significantly are rotavirus, coronavirus, cryptosporidium, K99+ Escherichia coli and Salmonella species, and the importance of these agents has been investigated recently by means of a survey of the microbiology of calf diarrhoea in southern Britain . During the survey, 21 moribund calves with diarrhoea were purchased and examined post mortem with the following objectives: to describe the pathology of the natural disease and to investigate diagnostic problems by studying the relationship between the lesions and infectious agents.
The lesions detected in eight of the 21 calves, which were infected with rotavirus, have been described in a preliminary report (Hall et al 1986) .
Twenty-one calves with diarrhoea were purchased from 11 of the 34 outbreaks surveyed (Table 1 ). The mean age of the calves was 11'4 days (range one to 23) and the mean duration of diarrhoea before necropsy was 2· 9 days (range O·5 to 14). Several breeds and crossbreeds were represented within the group which contained males and females in almost equal numbers. Some calves had been treated with antibiotics. Faeces were collected from all calves on the day of necropsy and from seven calves on one or two occasions during the week before necropsy. They were examined for rotavirus and coronavirus by enzyme-linked immunosorbent assay, for cryptosporidium by microscopical examination of faecal smears stained by the Giemsa method, for K99+ E coli by slide agglutination tests of 10colonies per faecal sample and for Salmonella species by enrichment of faeces in Rappaport and selenitebrilliant green broths and subculture to brilliant green agar (Jones et aI1983) . Fivegrams of faeces, prepared for electron microscopy by differential centrifugation followed by negative staining, were examined for Newbury agent (Bridger et al 1978). ...
Intestinal tissues for microscopy were removed under pentobarbitone sodium anaesthesia. Short lengths (10 em) of small intestine (51), referred to as duodenum and 511 to 519, and four lengths (5 ern) of spiral colon were ligated and filled with fixative; fixatives used were mercuric formol, a fixative comprising nine parts saturated mercuric chloride and one part glacial acetic acid, 12 per cent neutral buffered formalin and 0·1 M phosphate buffered 3 per cent glutaraldehyde, pH 7· 3. Duodenum was sampled adjacent to the pylorus, 511 close to the ligament of Treitz and 519 adjacent to the ileocaecal junction. Sites 512 to 518 were spaced as equally as possible between sites 1 and 9.
The calves were killed by intravascular injection of an overdose of anaesthetic and the sacs excised and, 240 (Table 2) . After fixation for 24 hours the mercuric formol and the mercuric chloride and acetic acid mixture were replaced with 80 per cent alcohol.
The immunostaining methods used were an indirect immunoperoxidase method with primary antisera to bovine rotavirus, bovine coronavirus, Newbury agent SRV-J and purified K99adhesin prepared in calves and rabbits and the peroxidase-antiperoxidase method (Sternberger et al 1970) with primary antisera to atypical E coli (SI02-9) .To obtain the enteropathogen scores the extent of infection of the cytoplasm or surface of enterocytes in the small and large intestines by enteropathogenic viruses, bacteria (salmonellae excluded) or protozoa, seen in a histological cross-section of intestine was assessed subjectively and scored as follows: o no enteropathogen-infected enterocytes detected; I -at least one enteropathogen-infected enterocyte detected; 2 -scattered enteropathogen-infected enterocytes detected;
• 3 -many enteropathogen-infected enterocytes detected; 4 -majority of enterocytes infected with enteropathogens; and 5 -all enterocytes infected with enteropathogens.
The sum of the scores of enteropathogens detected at a site was divided by five (the number of enteropathogens scored) to give a mean enteropathogen score for each site. Calves infected with Salmonella typhimurium were excluded from the calculations because in salmonellosis, surface infection comprises . only part ofthe disease process and severe lesions may result from infection in the lamina propria.
To obtain lesion scores, one section from each of the 10small intestinal sites was scored on a scale from o to 5 for the following 11 lesions: villus stunting, villus fusion, presence of degenerate cells in the lamina propria, presence of leucocytes in lacteals, cuboidal enterocytes, exfoliated enterocytes, vacuo- lated enterocytes, abnormally arranged enterocytes and presence of neutrophilic polymorphonuclear leucocytes in the lamina propria, crypts and intestinal lumen. One section of caecum, colon and rectum was scored similarly for the following five lesions: exfoliated enterocytes, abnormally arranged enterocytes and presence of neutrophils in the lamina propria, crypts and intestinal lumen. A maximum score of 55 could be allocated to each small intestinal site and 25 to each site in the large intestine. Scores at 10 sites in the small intestines of IS calves, nine sites from three calves, eight sites from one calf and six sites from two calves were pooled to derive the contribution of each lesion type to the overall pathology. Similarly, in the large intestines, scores from three sites from 17 calves, two sites from three calves and one site from one calf were pooled. Intestinal tissues which were not scored and non-intestinal tissues were examined and abnormalities which were detected were recorded.
To gain an overall impression of which parts of the intestinal tract were most severely damaged, the sum of the lesion scores at each site was divided by the number of calves in which that site was examined, to give a mean lesion score; the mean lesion score at each site was also calculated, excluding the three salmonella-infected calves.
For statistical analysis mean lesion scores and the mean of all enteropathogen scores were calculated for upper (511 to 513), middle (514 to 516) and lower (517 to 519) small intestine and large intestine (caecum, colon and rectum); salmonella-infected calves were excluded from the analysis. Correlations were calculated between the two scores at each region.
When histological examination revealed the presence of a lesion or infection which required further investigation, intestinal tissues fixed in glutaraldehyde were examined by scanning electron microscopy and transmission electron microscopy. Pieces of intestine were stored in the original glutaraldehyde at 4°C until required and processed for examination in transmission and scanning electron microscopes (Hall et al 1976 .
Faeces were not examined ante mortem for Newbury agent or for bacteria apparently adherent to the mucosal surface of the large intestine, and detections of these agents have beefi excluded from comparisons of results 'of faecal and mucosal examinations. Detection of enteropathogens by examination of faeces or the intestinal mucosa gave identical results in nine of the calves (2, 3, 5, 8, IS, 16, 17, 19 and 21) (Table 1) . More enteropathogens were detected by mucosal examination than by faecal examination in nine calves (4,6,7,9, 10, 11, 12, 14 and 20) ; the opposite was true for two calves (I and 13) and in another (18) rotavirus was detected in the faeces and coronavirus in the mucosa. Combined faecal and mucosal examinations detected four enteropathogens in two calves, three in nine calves, two in seven, a single enteropathogen (rotavirus) in one calf and no enteropathogens in another.
In one calf (3) of 13 rotavirus-infected calves, rota- in natural (Langpap et al 1979) and experimental (Mebus et al 1973) infections. In the nine calves remaining, coronavirus was either absent from the small intestines (five calves), or coronavirus scores in the small intestines were low (four calves). In these nine calves other enteropathogens were detected which probably contributed to the disease process; one additional enteropathogen was detected by mucosal examination in two of the nine calves, two additional enteropathogens in five calves, and three additional enteropathogens in two of the nine calves. Bacteria were detected apparently adherent to the surface of the large intestine in seven calves. Cryptosporidium was detected by mucosal examination in seven of the nine, rotavirus in three and S typhimurium in one.
,.
..' ....
If", -' ", . ,.;. , . ., ..,.' " virus was the only enteropathogen detected by mucosal examination (Table I) . High rotavirus scores were recorded throughout the small intestine and infection was present in the large intestine (Table 3) ; in this calf diarrhoea appeared to have been caused by rotavirus infection alone. In five of the remaining 12 rotavirus-infected calves, one additional enteropathogen was detected by mucosal examination, in four calves two additional enteropathogens and in three calves three additional enteropathogens were detected. Bacteria, apparently adherent to the surface of the large intestine, were seen in nine rotavirusinfected calves, coronavirus in five, cryptosporidium in six and Newbury agent in two. In three rotavirusinfected calves (2, 5 and 7), despite the presence of other enteropathogens, the severity (Fig 1) and extent (Table 3 ) of rotavirus infection in the small intestine suggested that rotavirus was the important aetiological agent. Lesions detected in the small intestines of calves in which rotavirus appeared the sole or principal cause of diarrhoea (2, 3, 5 and 7) were identical to those described in natural (Pearson et al 1978 and experimental (Mebus and Newman 1977) infections. Severe infection and lesions were distributed widely in the small intestines of these calves (Fig I, Table 3 ).
Cryptosporidium was detected in 11 calves by either faecal or mucosal examination, by both methods in four calves, and by mucosal examination alone in seven calves. In six calves (4, 12, 14, 16, 18 and 20) mucosal examination showed infection with cryptosporidium to be severe and extensively distributed in mid and lower small intestines ( Table 3 ), suggesting that it was the principal cause of diarrhoea. Infection with cryptosporidium was restricted largely to the small intestines of these six calves and the lesions were similar to those described previously , Pearson et al 1982 , Sanford and Josephson 1982 . In all 11 calves other enteropathogens, which could have contributed to the disease process, were detected by mucosa] examination. One additional enteropathogen was Cletected by mucosal examination in two of the 11 calves, two additional enteropathogens in seven calves and three additional enteropathogens in two calves. Bacteria were detected apparently adherent to the surface of the large intestines of eight of the I 1 cryptosporidium-infected calves. Coronavirus was detected by mucosal examination in seven of the II, rotavirus in four Newbury agent in two and S typhimurium in one.
Coronavirus was detected in 10 calves by either faecal or mucosal examination; by both methods in three calves, and by mucosal examination alone in seven calves. In one calf (13) high coronavirus scores were present throughout the small and large intestines and coronavirus was the only enteropathogen detected by mucosal examination; in this calf, diarrhoea appeared to have been caused by coronavirus infection. The lesions in the small and large intestines of this calf were identical to those described
Bacteria, apparently adherent to the surface of the large intestines, were detected by mucosal examination in 13 calves, always in association with one or more additional enteropathogens which could have contributed to the disease process. Faeces were not examined ante mortem for these bacteria because their presence was not suspected until mucosae were examined post mortem. One additional enteropathogen was detected by mucosal examination in four of the 13 calves, two additional enteropathogens in seven of the 13 calves and three additional entero- It 3 10 17 17 3 3 12 0 1 10 8 (17) or K99+ E coli (19) . In two calves (18 and 21), there was disarrangement of enterocytes and some had' exfoliated; enteropathogens were not detected.
Enterocytes infected with Newbury agent were detected by mucosal examination, in the intestines of three calves (7, 18 and 19); additional pathogens which could have contributed to the disease process were also detected. Cryptosporidium was detected in one calf (18) and K99+ E coli in another (19). In the third calf (7), infection with Newbury agent extended throughout the small and large intestine but it was not possible to conclude that Newbury agent was a causal agent because the calf was infected extensively with rotavirus and cryptosporidium. Nevertheless, lesions in the small intestines of the three calves were identical to those described in calves infected experimentally with Newbury agent (Hall et alI984) . pathogens in two calves. Rotavirus was detected in eight calves, coronavirus in seven, cryptosporidium in eight and S typhimurium in one calf. In two calves (6 and 9), the surface of the large intestine was infected severely and the severity of infection of the small and large intestines by other enteropathogens was slight; in these calves the adherent bacteria appeared to be the principal cause of diarrhoea. Villi in the small intestines of these calves (6 and 9) were slightly stunted and some were fused, but these lesions could have been caused by enteropathogenic viruses which were detected in these calves. Lesions seen in the large intestines were disarrangement and exfoliation of surface enterocytes and the lamina propria was infiltrated by neutrophils. Light and electron microscopic examination of the mucosal surfaces of the large intestines of the 13 calves, in which bacteria were visible adherent to the mucosal surface, revealed variation in the nature of the bacterial colonisation and in the severity of the lesions. In seven calves (1, 6, 9, 10, 11, 12 and 15) , electron microscopic examination revealed that the bacteria were closelyattached to the surface of enterocytes (Fig 2) , often by means of 'cups' or 'pedestals', and microvilli were effaced (Fig 3) . In the remaining six calves (2, 4, 5, 8, 14 and 16) the bacteria appeared, by light microscopy, to be adherent to the mucosal surface but examination by transmission electron microscopy revealed that, although closely associated with the musocal surface, they were not attached to the surface of enterocytes; they were detected most frequently in folds and crevices in the mucosa (Fig 4) . In some calves, bacteria formed part of an exudate together with mucus and exfoliated cells. The lesions detected in both groups were similar and comprised disarranged enterocytes, exfoliated enterocytes ( Fig 5) and infiltration of the lamina propria by neutrophils; lesions varied in severity. Scanning electron microscopy revealed that in some animals bacteria were associated closely with enterocytes which were disarranged and exfoliated, but in other animals bacteria were not adherent to the abnormal enterocytes although bacteria were associated with adjacent cells. Histological sections from the large intestines of the 13 calves in which bacteria were visible apparently adherent to the mucosal surface, were stained by the immunoperoxidase method , with a primary antiserum raised against E coli (SI02-9) which had been identified as a cause of dysentery in neonatal calves . In seven calves in which there were lesions identical to those caused by S102-9the bacteria were stained in fivecalves (1,9,10, 11 and 15) but not in two calves (6 and 12).
Lesions were not detected in the large intestines of two of the eight calves in which bacteria were not associated with the mucosal surface of the large intestines (3 and 7) ; the lesions detected in four of the
Lesions were detected in the small intestines of four calves (8, 10, 11 and 15) in which no one enteropathogen appeared to play a major role; several enteropathogens were detected and bacteria, apparently adherenl to the surface of the large intestines, were present. The lesions in these calves comprised stunted and fused villi covered by dis- Escherichia coli K99+ E coli were detected, by faecal and mucosal examination, in one calf (19). Newbury agent was detected also in the small intestinal mucosa of this calf, but the severity and extent of infection of the small intestinal surface by E coli suggested that this agent was the principal cause of diarrhoea. The lesions in the small intestine of this calf comprised mild enterocyte exfoliation and slight stunting and fusion of villi, which could have been caused by Newbury agent (Hall et al 1984) , and infiltration of Salmonella typhimurium S typhimurium was detected in three calves by either faecal or mucosal examination, by both methods in two calves and by faecal examination alone in one. Mucosal examination detected other enteropathogens in three calves; coronavirus in one (17), cryptosporidium and bacteria apparently adherent to the surface of the colon in another (16) and rotavirus and E coli (SI02-9) adherent to the surface of the small and large intestine in a third (1). The extent of infection and the nature of the lesions in two calves (1 and 17) suggested that S typhimurium was the principal cause of diarrhoea. Villi in the small intestine of these calves were stunted and fused, enterocytes had exfoliated and remaining enterocytes were cuboidal and disarranged. Neutrophils were numerous in the lamina propria, crypts (numerous crypt abscesses were seen) and intestinal lumen. arranged and cuboidal enterocytes; exfoliated enterocytes were present. In the large intestines of these calves there were disarranged and exfoliated enterocytes on the mucosal surface and mild infiltration of the lamina propria and crypts by neutrophils.
Mean enteropathogen scores (Fig 6) increased along the small intestine and were highest in the colon. The reason for high scores in the lower small intestine and the large intestine was the presence of multiple infections; coronavirus, rotavirus and cryptosporidium in the small intestine, and coronavirus and bacteria apparently adherent to the mucosal surface in the large intestine. The duodenum appeared resistant to infection by all enteropathogens except coronavirus, which was detected in the duodenum of three calves (10, 13 and 15).
Lesions were detected in the small intestines of all calves (Table 3 ) and scores at sites 5 to 9 were higher than those at sites 1 to 4 (Fig 6) ; lesions were detected rarely in the duodenum. Lesions were detected in the large intestines of 19 of the 21 calves (Table 3 ) and in nine of these calves (5, 6,8,9,12,13,15,16 and 17) severe lesions were seen (Table 3 ). The contribution of the 11 lesions scored in the nine small intestinal sites, as a percentage of the overall pathology was: stunted villi 23 per cent, fused villi 18 per cent, cuboidal enterocytes 15 per cent, disarranged enterocytes 12 per cent, exfoliated enterocytes 12 per cent, neutrophils in the lamina propria 6 per cent, vacuolated enterocytes 5 per cent, neutrophils in crypts 4 per cent, neutrophils in the gut lumen 2 per cent, degenerate cells in the lamina propria 2 per cent and cells in lacteals 1 per cent. The contribution of the five lesions scored in the large intestine as a percentage of the overall pathology was: exfoliated enterocytes 35 per cent, disarranged enterocytes 30 per cent, neutrophils in the lamina propria 20 per cent, neutrophils in crypts 9 per cent and neutrophils in the lumen 6 per cent.
Lesion scores and pathogen scores were correlated significantly. Examination of each region separately indicated correlations between O'53 and O'59, which were significant statistically; P
|
