Study on the Epidemiology and Control of Campylobacter

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1992, p /92/ $02.00/0 Copyright ( 1992, American Society for Microbiology Vol. 58, No. 6 Study on the Epidemiology and Control of Campylobacter jejuni in Poultry Broiler Flocks ARJEN VAN DE GIESSEN,"* SYLVIE-ISABELLE MAZURIER,2 WILMA JACOBS-REITSMA,3 WIM JANSEN,4 PETER BERKERS,' WILMA RITMEESTER,' AND KAREL WERNARS1 Laboratory for Water and Food Microbiology' and Laboratory for Bacteriology, 4 National Institute of Public Health and Environmental Protection, 3720 BA Bilthoven, The Netherlands; Laboratoire de Genie des procedes et d 'hygiene alimentaire, Institut National de la Recherche Agronomique, Massy, France2; and Spelderholt Centre for Poultry Research and Information Services, Agricultural Research Service (DLO), 7361 DA Beekbergen, The Netherlands3 Received 27 December 1991/Accepted 26 March 1992 Broiler flocks are frequently infected with Campylobacterjejuni. The origin of the infection is still unclear. The question of whether colonization of flocks results from transmission of C. jejuni from breeder flocks to progeny (vertical transmission) or from environmental sources (horizontal transmission) remains to be answered. Therefore, in this study samples were taken from successive broiler flocks in two broiler houses (house A on farm A and house Bi on farm B) as well as from the environment of the houses. All C.jejuni isolates were typed by using the Penner serotyping system, and part of the isolates from farm B were typed by using a randomly amplified polymorphic DNA-typing system. In poultry house A, C. jejuni was isolated from the first flock but not from subsequent flocks. In poultry house B1, C. jejuni strains of the same Penner serotypes and exhibiting identical DNA profiles were isolated from successive flocks. Infection of the flocks from a common source via horizontal pathways is suspected, while a vertical route of infection is not likely to exist. Application of measures to control horizontal transmission of C. jejuni on farm B was successful. During the past decade, Campylobacter jejuni has been recognized as a major cause of human gastroenteritis. In several developed countries the incidence of campylobacteriosis exceeds even that of salmonellosis (6, 16). In the Netherlands, sentinel studies revealed that campylobacteriosis accounts for approximately 12% of all cases of acute human gastroenteritis (8). Foods of animal origin may serve as vehicles of infection (3). In many reports, an association has been made between campylobacteriosis and recent consumption of raw or undercooked poultry meat (5, 12). The percentage of infected broiler chickens at slaughter is frequently high. During the slaughtering process C. jejuni can easily spread from the intestinal contents to the carcasses (13). Several studies of the prevalence of C. jejuni in fresh and frozen poultry meat yielded a high percentage of contamination (1, 13). Consequently, this bacterium has become a major concern to poultry industry. Up till now, however, the pathways involved in the infection of poultry flocks are still unclear. Several factors suspected to be sources or vectors of infection have been the subjects of studies. Kazwala et al. (9) suggested that flocks were most likely infected from the environment of the poultry houses. Genigeorgis et al. (7) suspected transmission of C. jejuni from one generation of chickens to the next via the old litter. However, Lindblom et al. (10) found that environmental samples from poultry houses, taken before colonization of the birds, were always negative. Pearson et al. (14) isolated C. jejuni from small animals (not further identified) on the farm but not from feed, litter, air, shed walls, or floors. In the same study, C. jejuni was detected by direct immunofluorescence microscopy in the water of the supply system coming from a bore hole. Other possible sources include personnel, dogs, cats, flies, and rodents (3). Although C. jejuni may be * Corresponding author occasionally present on the shells of freshly laid eggs (4), the organism was not isolated from day-old chicks and a vertical route of transmission from infected breeder flocks to progeny is doubted by several authors (2, 4, 7). However, newly hatched chicks housed in the protective environment of a laboratory still became colonized (10). The present longitudinal study was done in 1990 by using the Penner serotyping system as well as a DNA-typing system in order to assess the eventual roles of vertical and horizontal transmission routes in the infection of poultry flocks with C. jejuni. MATERIALS AND METHODS Farms. During 1990 and the first part of 1991 a longitudinal study was carried out involving two broiler farms. On farm A only one broiler house, house A, was present. Cleaning and disinfection of this broiler house was carried out as a matter of routine between all successive broiler cycles, that is, after slaughtering of a broiler flock at an age of 6 weeks and before introduction of a new generation of 1-day-old chicks. Hygienic precautions included separate clothes and boots at the entrance of the house but did not include a footbath disinfection barrier. Rodents, including rats and mice, and insects were controlled chemically. Chicks were supplied with tap water and pelletized feed, and wood shavings were used for litter. Apart from poultry, 50 pigs were present on this farm and kept in a pig house. A cat and a dog were present as well. On farm B three broiler houses were present: houses Bi, B2, and B3. The all-in all-out system was used, which means that broiler flocks in these houses were removed and slaughtered simultaneously and new flocks of 1-day-old chicks were brought in simultaneously as well. Between successive cycles the houses were cleaned with water obtained from a pool in the direct vicinity of the farm. The houses were not disinfected. A lot of rubbish was present on this farm in the surroundings of the broiler houses, and no hygienic precau-

2 1914 VAN DLE GIESSEN ET AL. tions were taken at the entrance of the houses. Insects and rodents, including mice and rabbits, were present on a large scale and were not under control. Chicks were supplied with tap water and pelletized feed. Chopped straw was used as litter, and stock was stored in an open shed accessible to vermin. Twenty pigs, kept in a pig house, as well as two cats and two dogs were present on this farm. Now and then, the dogs were fed with dead chicks from broiler house Bi. On both farms the broiler houses were kept empty between successive flocks for approximately 1 week. Sampling. Seven successive broiler flocks housed in house A and ten successive flocks housed in house Bi were sampled. Each flock in house A consisted of 15,000 birds, and each flock in house Bi consisted of 30,000 birds. The origin of the flocks was not known. In house A at least four different breeds and in house Bi at least two different breeds were involved. All flocks were sampled at an age of 4 to 5 weeks by using sterile cotton swabs. Each flock was sampled by taking 30 swabs at random of fresh cecal droppings. Swabs were put into tubes containing 2 ml of phosphatebuffered saline and transported to the laboratory for microbiological examination. All samples were examined for the presence of Campylobacter spp. If Campylobacter spp. were not detected in the fecal samples of a flock, ceca from 30 birds of the flock were collected at the slaughterhouse and cecal contents were examined for the presence of Campylobacter spp. On farm A, 19 and 20 swabs of fresh fecal material were taken from the pigs during broiler cycles 5 and 7, respectively. On farm B broiler flocks present in houses B2 and B3 were sampled during cycle 5. These flocks consisted of 30,000 birds each. From these flocks only five swabs were taken. From the pigs on farm B one and three swabs of fresh fecal material were taken during cycles 7 and 8, respectively. During cycle 7 one sample each of litter, feed, and drinking water were taken from house Bi. At the same time several samples were taken from the environment of this broiler house, including two rectal swabs from a dog and one from a cat, one swab of fresh fecal material from a dog and one swab of fresh fecal material from rabbits, one swab from the cecal content of a mouse, one sample of chopped straw from the open shed, and one sample of surface water from the pool. All samples were examined for the presence of Campylobacter spp. Isolation and serotyping of Campylobacter spp. Portions of 1 g of litter, feed, and straw and 1 ml of drinking and surface water were added to 20 ml of Campylobacter selective enrichment broth (THAL) (thioglycolate broth [BBL 11260] with 5% horse blood, vancomycin [0.04 g/liter], polymyxin B sulfate [0.01 g/liter], trimethoprim [0.02 g/liter], cycloheximide [0.1 g/liter], and cephalotin [0.1 g/liter]). After incubation at 37 C for 24 h in a microaerobic atmosphere (7% 02, 10% C02, and 83% N2), the broth was streaked onto Campylobacter selective blood-free agar (CCD agar [Oxoid CM 739] with cefoperazone [32 mg/liter] and cycloheximide [100 mg/liter]). Swab samples were plated out directly by swabbing the surface of the CCD agar. Plates were incubated at 37 C for 48 h in a microaerobic atmosphere. Characteristic colonies were selected and examined under a phase-contrast microscope for typical spiral-shaped cells and rapid motility. From each plate one colony suspected of being a Campylobacter spp. was transferred onto CCD agar. After another incubation period under the same conditions, isolates were confirmed biochemically by the following tests (criteria for C. jejuni in parentheses): catalase (positive), oxidase (positive), growth on blood agar at 43 C after 2 days (positive) and at 25 C after 5 days (negative), growth in brain heart infusion broth containing 3.5% NaCl after 2 days (negative), reduction of nitrate (positive), hydrolysis of hippurate (positive), and sensitivity to nalidixic acid (positive). All isolates were serotyped by using the method described by Penner and Hennessy (15). RAPD analysis. Several C. jejuni isolates from the flocks and the environment of farm B were typed by using a novel technique called randomly amplified polymorphic DNA (RAPD) analysis. This method is based on the in vitro amplification of polymorphic DNA sequences by using a polymerase chain reaction format. The method was carried out as described by Mazurier et al. (11). In summary, C. jejuni colonies were inoculated in brain heart infusion broth and grown microacrobically for 18 h at 37 C under vigorous agitation. One milliliter of the culture was centrifuged, and the cell pellet was resuspended in 1 ml of distilled water. Then the suspension was heated to 100 C for 10 min and centrifuged. Five microliters of the supernatant was used in the amplification reaction. In this reaction one primer, HLWL85, was used. After amplification, the reaction mixture was electrophoresed on a 1.6% agarose gel, stained with ethidium bromide, and photographed under UV transillumination to visualize the RAPD DNA profile. Preventive measures taken at farm B. After broiler cycle 7 on farm B, preventive measures were taken by the farmer to prevent Campylobacter infection of the broiler flocks. These measures included cleaning of the broiler houses with clean tap water and disinfection with Formalin at a concentration of 20% between successive cycles. To make cleaning and disinfection more effective, the floors of the broiler houses were leveled with concrete and the walls were made as smooth as possible. Instead of the chopped straw from the open shed, squeezed wood shavings were used as litter. At the entrance of the broiler houses, hygienic precautions were taken including a hygiene barrier, separate boots and clothes, footbath disinfection, and hand-washing facilities. The farm yard was cleaned up and coated with concrete, and a gate was placed around it. The broiler houses were made inaccessible to vermin, and vermin were controlled chemically. RESULTS APPL. ENVIRON. MICROBIOL. Farm A. In broiler house A, C. jejuni was detected in flock 1 but not in flocks 2 through 7 (Table 1). In flock 1, C. jejuni was isolated from 20 of 30 samples of feces, including four different serotypes. C. jejuni was also detected in the pigs present on this farm (Table 1). During broiler cycle 5, C. jejuni was isolated from 3 of 19 samples of pig feces, including two different serotypes. During cycle 7, C. jejuni was isolated from 12 of 20 samples of pig feces, including seven different serotypes. Serotype 049 was isolated from both the pigs and the first broiler flock. Farm B. In broiler house Bi, C. jejuni was isolated from the majority of the fecal samples from flocks 1, 2, 4, 5, 6, and 7 (Table 2). C. jejuni was also isolated from 27 of 30 cecal samples from flock 3 at slaughter. Only four different serotypes were distinguished. Serotype 059 was frequently isolated from flocks 1, 4, 5, 6, and 7 but was not isolated from flocks 2 and 3. RAPD typing of 059 isolates from flocks 1, 4, 5, 6, and 7 yielded identical DNA profiles (indicated as type A). Also, isolates from flocks 1 and 7 that could not be serotyped were of RAPD type A. Serotype 042 was frequently isolated from flocks 2 (all 30 samplcs), 3, and 6 but was not isolated from flocks 4, 5, and isolates from flocks 2, 3, and 6 exhibited identical DNA profiles (indicated

3 VOL. 58, 1992 TABLE 1. Occurrence of C. jejuni in successive broiler flocks in broiler house Al and in pigs on farm A Broilers Pigs Broiler cycle no. Penner No. of Penner No. of serotype samples serotype samples NDa 10 2 ND 30 3 ND 30 4 ND 30 5 ND U* 1 ND 16 6 ND 30 7 ND U 2 ND 8 a ND, Campylobacter spp. were not detected. " U, untypeable. as type B). Serotype 053 was isolated from a single sample from flock 3. The DNA profile of this isolate was indicated as type C. Two isolates from flock 4 that could not be serotyped were of RAPD type C as well. Serotype 05 was isolated from three samples of flock 7. One 05 isolate was RAPD TABLE 2. Occurrence of C. jejuni in successive broiler flocks in broiler house Bi on farm B Cycleno. Penner RAPD type No. of Cycle no. serotype (no. of isolates) samples A (1) 9 ua A (2) 5 NDb B (2) B (2) C(1) 1 U NT' A (2) 14 U C (2) 12 ND A (2) 22 ND B (1) A (2) 19 U D (1) D (1) A (2) 21 U A (1) 1 ND 5 8 ND 30 9 ND ND 30 U, untypeable. a "ND, Campylobacter spp. were not detected. NT, not typed. C. JEJUNI IN POULTRY 1915 TABLE 3. Occurrence of C. jejlni in materials from house Bi and in the environment of farm B including houses B2 and B3 Material or Broiler Penner RAPD type No. of animal cycle no. serotype (no. of isolates) samples Poultry House B A (1), E (1) 3 U" A (2) 2 House B A (2) 3 U A (2) 2 Litter 7 NDh 1 Feed 7 ND 1 Drinking water 7 ND 1 Pigs F (1) G (1) H (1) 1 U I (1) 1 Rabbits J (1) 1 Mouse 7 ND 1 Cat 7 ND 1 Dogs 7 ND 3 Water pool 7 ND 1 Straw (shed) 7 ND 1 " U, untypeable. " ND, Campylobacter spp. were not detected. typed and appeared to be of the same RAPD type (indicated as type D) as one isolate from flock 6 that could not be serotyped. The results of the other samples taken from broiler house Bi and from the environment are summarized in Table 3. Campylobacter spp. were not detected in samples of litter, feed, and drinking water taken from house B1. C. jejuni was isolated from both broiler flocks present in the adjacent broiler houses B2 and B3 during broiler cycle 5. In both flocks serotype 059 was detected. 059 isolates from these flocks, as well as isolates from these flocks that could not be serotyped, were shown to be of RAPD type A. However, one 059 isolate from the flock in house B2 appeared to be of a different DNA type (indicated as type E). Also, C. jejuni was isolated from the pigs present on this farm, including four different serotypes and four different DNA types (indicated as types F, G, H, and I). From the fecal material of rabbits a C. jejuni isolate of serotype 019 was obtained. The DNA type of this isolate was indicated as type J. Campylobacter spp. were not detected in the samples taken from a mouse, a cat, dogs, the water pool, and straw from the shed. In pursuance of the results described above, the owner of farm B was willing to adjust his farm management in order to control Campylobacter infection of subsequent broiler flocks. Campylobacter spp. were not isolated from later broiler flocks (numbers 8, 9, and 10) in house Bi. DISCUSSION In house A, broiler flock 1 was found C. jejuni positive, whereas the subsequent six flocks were not infected with this organism. Transmission of C. jejuni from the first flock to the following flocks did not occur. If C. jejuni is able to persist in poultry houses for several days after depopulation, transmission of the organism to the following flocks in broiler house A has probably been prevented by efficient cleaning and disinfection. Remarkably, transmission of C. jejuni from the infected pigs to the broiler flocks 2 through 7 did not occur. This may be due to the separate clothes and boots used for the broiler house, although no further hygienic precautions

4 1916 VAN DE GIESSEN ET AL. were taken. From the results of house A it can also be concluded that either vertical transmission of C. jejuni is improbable or the mother flocks delivering the chicks in cycles 2 up to 7 happened to be free of C. jejuni. However, the different breeds involved indicate that at least four mother flocks were involved. Moreover, mixing of eggs at the hatcheries has become common practice, resulting in broiler flocks descending from more than one mother flock. Regarding this fact and the presumed high incidence of C. jejuni in poultry flocks in the Netherlands (18), the chance that all mother flocks involved in this case were free of C. jejuni is quite small. Therefore, this result suggests that a vertical infection route is not likely to exist. A completely different picture is obtained from the results of farm B. C. jejuni was detected in seven successive flocks in broiler house Bi. Serotypes 059 and 042 were isolated from several of the successive flocks. RAPD typing of 059 isolates yielded identical DNA profiles. Further, 042 isolates from different flocks appeared to be of the same RAPD type. A third type of C. jejuni was found in flocks 3 and 4, and a fourth type showed up in flocks 6 and 7. In summary, the results of both the serotyping and the RAPD typing suggest that the same strains of C. jejuni were involved in successive flocks. In this case, infection of the flocks from a common source is probable. A vertical route of infection from infected breeder flocks to the broiler flocks cannot be excluded. However, after introduction of control measures C. jejuni was no longer detected. Therefore, C. jejuni was probably transmitted to the broiler flocks by horizontal routes. The source of infection, however, is not clear. Isolation of the C. jejuni from broiler flocks in houses Bi, B2, and B3 suggests that either cross-contamination did occur or the broiler flocks were infected from the same environmental source. The C. jejuni-positive pigs may have been the source of infection. However, the few strains isolated from the pigs differed from the poultry isolates in serotype or RAPD type. Also, C. jejuni-positive rabbits may have been the source of infection on this farm, but only a single isolate was obtained from these animals and this isolate was of a different serotype and RAPD type. Other possible sources in the environment of the farm were all found to be Campylobacter sp. negative. Finally, broiler house Bi itself may have been the source of infection. Persistence of C. jejuni in the house between successive flocks may well have been facilitated by the lack of disinfection on this farm. In this case, however, the question is why the serotypes skipped some cycles, that is, why they were not isolated from all successive flocks. In all probability, strains of C. jejuni introduced into a poultry flock may not have equal opportunities for growth and spread. Certain types may easily emerge and be frequently isolated, whereas others may become overgrown and not be isolated at all. If successive flocks are infected with the same C. jejuni strains, emergence of different strains per flock may result. The persistence of C. jejuni in niches of the broiler house and alternate emerging of serotypes in successive flocks may well be the best explanation for the results of poultry house Bi. At last, control measures, including disinfection of the broiler houses and hygienic barriers, have been shown very effective in eliminating C. jejuni from broiler house Bi and preventing reintroduction of this organism from the environment. C. jejuni contamination of poultry meat as a consequence of infected poultry flocks has become a serious problem for the poultry industry. Although the organism is of little veterinary concern, having a nonpathologic, commensal relationship with the chicken (17), the role of poultry meat in causing human campylobacteriosis cannot be denied. Despite the reduction in C. jejuni during processing, the organism may still be present on a large percentage of poultry products, even in large numbers (13). However, C. jejuni is not part of the normal chicken flora and some broiler flocks appear to stay free of this organism. Therefore, strategies to prevent contamination of poultry meat should emphasize the Campylobacter-free raising of the flocks in the first place. The results of this study indicate that C. jejuni contamination of broiler flocks may occur by horizontal pathways, whereas a vertical route of infection is not likely to exist. Moreover, the results of this study show that control of C. jejuni by efficient cleaning and disinfection of broiler houses combined with strict hygiene can be effective. However, the epidemiology of C. jejuni in poultry flocks should be elucidated in more detail to provide a basis for a more specific control strategy. For this, application of the RAPD DNA technique has been shown to be very useful in this study. ACKNOWLEDGMENTS APPL. ENVIRON. MICROBIOL. This study was carried out on behalf of the Dutch Public Health Veterinary Inspectorate (VHI). We thank M. Hout (VHI) for sampling. REFERENCES 1. Aho, M., and J. Hirn Prevalence of campylobacteria in the Finnish broiler chicken chain from the producer to the consumer. Acta Vet. Scand. 29: Annan-Prah, A., and M. Janc The mode of spread of Campylobacter jejuni/coli to broiler flocks. J. Vet. Med. B 35: Blaser, M. J., D. N. Taylor, and R. A. Feldman Epidemiology of Campylobacter jejuni infections. Epidemiol. Rev. 5: Doyle, M. P Association of Campylobacter jejuni with laying hens and eggs. Appl. Environ. Microbiol. 47: El-Shenawy, M. A., and E. H. Marth Campylobacter jejuni and foodborne campylobacteriosis: a review. Egypt. J. Dairy Sci. 17: Foster, E. M New bacteria in the news. A special symposium campylobacteria jejuni. Food Technol. 40: Genigeorgis, C., M. Hassuney, and P. Collins Campylobacterjejuni infection on poultry farms and its effect on poultry meat contamination during slaughtering. J. Food Prot. 49: Hoogenboom-Verdegaal, A. M. M., M. During, A. Leentvaar- Kuypers, P. G. H. Peerbooms, W. C. M. Kooij, R. van Vlerken, and H. Sobczak Epidemiologisch en microbiologisch onderzoek met betrekking tot gastro-enteritis bij de mens in de regio's Amsterdam en Helmond, in In RIVM report no National Institute of Public Health and Environmental Protection, Bilthoven, The Netherlands. 9. Kazwala, R. R., J. D. Collins, J. Hannan, R. A. P. Crinion, and H. O'Mahony Factors responsible for the introduction and spread of Campylobacter jejuni infection in commercial poultry production. Vet. Rec. 126: Lindblom, G.-B., E. Sjogren, and B. Kaijser Natural Campylobacter colonization in chickens raised under different environmental conditions. J. Hyg. Camb. 96: Mazurier, S.-I., A. W. van de Giessen, K. Heuvelman, and K. Wernars. Lett. Appl. Microbiol., in press. 12. Oosterom, J., C. H. den Uyl, J. R. J. Banifer, and J. Huisman Epidemiological investigations on Campylobacter jejuni in households with a primary infection. J. Hyg. Camb. 92: Oosterom, J., S. Notermans, H. Karman, and G. B. Engels

5 VOL. 58, 1992 C. JEJUNI IN POULTRY 1917 Origin and prevalence of Camnpylobacterjejulni in poultry processing. J. Food Prot. 46: Pearson, A. D., R. R. Colwell, D. M. Rollins, M. L. Hanninen, M. W. Jones, T. D. Healing, M. Greenwood, M. Hood, M. Shahamat, E. Jump, and D. M. Jones Transmission of Campylobacter on a poultry farm. In B. Kaijser and E. Falsen (ed.), Proceedings of the International Conference on Campylobacter. Goteborg, Sweden. 15. Penner, J. L., and J. N. Hennessy Passive hemagglutination technique for serotyping Campylobacterfetrns subsp. jejulni on the basis of soluble heat-stabile antigens. J. Clin. Microbiol. 12: Skirrow, M. B Campylobacter. Lancet 336: Stern, J., and R. J. Meinersmann Potentials for colonization of Campylobacter jejuni in the chicken. J. Food Prot. 52: van de Giessen, A. W., et al. Unpublished data.