- Email: [email protected]
Temporal Distribution and Genetic Fingerprinting of Salmonella in Broiler Flocks from Southern Japan
MOLECULAR, CELLULAR, AND DEVELOPMENTAL BIOLOGY Temporal Distribution and Genetic Fingerprinting of Salmonella in Broiler Flocks from Southern Japan F. Shahada,*1 T. Chuma,* K. Okamoto,* and M. Sueyoshi† *Laboratory of Veterinary Public Health, Department of Veterinary Medicine, Faculty of Agriculture, Kagoshima University, 1 21-24 Korimoto, Kagoshima 890-0065, Japan; and †Laboratory of Veterinary Hygiene, Department of Veterinary Medicine, Faculty of Agriculture, Miyazaki University, 1-1 Gakuenkonohanadai, Miyazaki 899-2192, Japan
Key words: Salmonella, broiler flock, temporal distribution, serotype, genotype 2008 Poultry Science 87:968–972 doi:10.3382/ps.2007-00455
Meat processing plants are potential hubs in the detection of infectious disease agents including Salmonella. Determining the Salmonella status of broiler flocks during slaughter is necessary to direct interventions at high prevalence holdings. Establishing the status of Salmonella isolates has been useful as a way of following trends over time, and also an important way in identifying common isolates that cluster over time or space (Olsen et al., 2001). A temporal cluster is a group of isolates of a particular serovar that aggregate together in time. Thus, identifying clusters can provide information that may be used to establish possible determinants of the disease and methods that can be used for control and prevention (Tauxe, 1991). Currently, there is limited information pertaining to the contamination of national and regional poultry facilities by Salmonella. Being part of the veterinary infectious agents monitoring and surveillance system in Japan, we report for the first time the incidence, temporal distribution, and genetic fingerprints of Salmonella in broilers in the Southern prefecture of Japan’s main island. A systematic abattoir longitudinal study was conducted to estab-
INTRODUCTION Serotypes of non-Typhi Salmonella enterica (hereafter referred to as Salmonella) are important pathogenic organisms in humans and other animals (Khakhria et al., 1997). Subclinical infection is not uncommon, and many animals, particularly poultry, become asymptomatic carriers, intermittently shedding the bacteria into the environment for variable periods of time. The latter situation is particularly important in the transmission and maintenance of Salmonella (Humphrey et al., 1998). Nontyphoid human infections are most often acquired from contaminated foods of animal origin. Salmonella Infantis and Salmonella Typhimurium have been major causes of salmonellosis in humans in Japan until 1989 when Salmonella Enteritidis suddenly emerged and continued prevailing to date (Hamada et al., 2003a; Asai et al., 2007).
©2008 Poultry Science Association Inc. Received November 7, 2007. Accepted January 29, 2008. 1 Corresponding author: [email protected]
968
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
holdings (n = 9), positive-flocks from persistently infected holdings (n = 21), and holdings (n = 19) that showed fluctuations with alternating negative and positive flocks for variable time periods. Fourteen holdings (negative, n = 5 and positive, n = 9) were sampled once throughout the study period. Seasonality component was not observed, and salmonellae were found colonizing broiler ceca in warm and cold months. Predominant serovar was Salmonella Infantis (93.3%; n = 525). Macrorestriction fingerprints of Salmonella Infantis using XbaI presumed the isolates to be derived from a common parent. Enhanced discrimination by BlnI digestion produced 3 banding patterns that were closely related genetically and hence epidemiologically related. Such epidemiological information may enable producers to formulate effective control action plan tailored for individual holdings with special emphasis on biosecurity, hygiene, and pest control.
ABSTRACT During the 1998 to 2003 period, cecal contents of 4,024 broiler chickens from 252 flocks raised in 63 holdings were examined for Salmonella. The aims were to establish the actual status of the infection, its temporal distribution, prevalent serotype, and common genotype among broiler flocks brought at the slaughterhouse. Collected samples were preenriched in Hajna tetrathionate broth, and after 24 h of incubation, 10 L of the broth was streaked on selective Rambach agar plate. Suspected scarlet color colonies of Salmonella were cloned on nutrient agar, confirmed through biochemical tests and serotyped using O and H antigens. Pulsed field gel electrophoresis technique generated DNA fragments banding patterns and established their clonal relatedness. Salmonella was isolated from 563 (14%) samples in 179 (71%) flocks. The flock situation varied from Salmonella-negative
969
PREVALENCE AND GENETIC FINGERPRINTS OF SALMONELLA
lish the percentage of positive flocks and identify the temporal trends and seasonality of Salmonella infections in broiler chickens.
MATERIALS AND METHODS Sample Collection and Preparation
Bacterial Isolation Approximately 1 g of cecal contents was aseptically mixed with 5 mL of sterilized distilled water and homogenized by vortexing. Then, 1 mL of the suspension was preenriched in 5 mL of Hajna tetrathionate broth (Eiken Chemical Co. Ltd., Tokyo, Japan) and incubated in a water bath at 42°C. After 24 h of incubation, a loopful from each of the enriched broth was streaked onto plates of selective Rambach (Rambach, 1990) agar, and incubated at 37°C for 24 h.
Bacteria Identification Suspected scarlet color colonies were selected from each plate and cloned on nutrient agar slants. Salmonella were confirmed using Gram staining and biochemical tests including fermentation of glucose, lactose and sucrose, hydrogen sulfide production, citrate, lysine decarboxylation, methyl red, and indole tests. Gram-negative rods with typical biochemical profiles of Salmonella were serotyped for somatic antigen by slide agglutination test using Salmonella group O antisera (polyvalent O, O4, O7, and O9). Salmonella isolates were further serotyped for flagellar (H) antigens by the tube agglutination test according to the Kaufmann-White scheme.
Pulsed Field Gel Electrophoresis A random sample (n = 36) of predominant O7 serogroup was drawn and DNA analyzed by pulsed field gel electrophoresis (PFGE). From June to December 1998, 1 sample was selected each month. From 1999 through 2003, 1 sample was drawn after every 2 mo. To trace the genetic diversity, samples were selected from different farms. Plugs preparation, XbaI and BlnI digestion were
Year
n samples
n flocks
n positive samples (%)
1998 1999 2000 2001 2002 2003 Total
381 732 768 704 704 735 4,024
24 46 48 44 44 46 252
57 (15) 116 (15.8) 115 (15) 88 (12.5) 76 (10.8) 111 (15.1) 563 (14)
carried out according to the procedure described by Ribot et al. (2006). Electrophoresis conditions, staining of gels, and digital images acquisition were performed as previously described by Shahada et al. (2007).
Data Analysis Statistical analysis was performed in an Excel spreadsheet. The percentage of positive flocks was calculated, and differences in mean prevalence were evaluated by using Microsoft Excel Analyze-It statistical tests for the χ2 and ANOVA. The probability level of 0.05 was used throughout. Analysis and interpretation of PFGE TIFF images was carried out as previous explained (Shahada et al., 2007).
RESULTS Prevalence of Salmonella Seventy-one percent of broiler flocks tested Salmonellapositive, and Salmonella was detected in 563 (14 %) out of 4,024 samples (Table 1). Out of 563 positive samples, 525 (93.3%) were serogroup O7, 6 (1.1%) serogroup O9, 2 (0.4%) serogroup O4, 10 (1.8%) polyvalent O, and 20 (3.55%) were not typed. Determination of H antigens in serogroup O7 revealed phase I (r) antigens produced by motile Salmonella Infantis and, hence, formed the predominant serotype. Serogroups O4 and O9 comprise S. Typhimurium and Salmonella Enteritidis serotypes, respectively. Of 252 flocks examined, 171 were positive with a single Salmonella serotype, whereas 8 revealed mixed serotypes. A comparison of means showed there was no statistical difference in the mean prevalence between the flocks.
Seasonality Component Salmonella prevalence trends did not vary when the 12mo time periods (January to December) were compared. A similar trend was also depicted by the cumulative 5yr period from January 1999 through December 2003. Minimum and maximum annual flock prevalence ranged between 56.8 and 79.5% during 2002 and 2001, respectively, whereas the cumulative monthly prevalence ranged between 56.5 and 82.6% in July and October, respectively. As a result, monthly and annually flock preva-
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
Samples were selected from 2 flocks on each sampling day. Sampled flocks were selected at random among 5 to 6 flocks slaughtered daily at the plant. Sample collection was carried out systematically at 14-d intervals throughout the study period. A total of 4,024 samples derived from 252 flocks were collected following 127 visits to the plant. Seventy-four of these flocks from 43 holdings were sampled 1 to 4 times; 84 flocks from 12 holdings were sampled 5 to 9 times; and 94 flocks originating from 8 holdings were tested 10 to 14 times. All samples were taken using sterile techniques, placed in sterile sampling plastic bags, and chilled by ice blocks during transportation. The samples were cultured on the same day upon arrival at the laboratory.
Table 1. Status of Salmonella in broiler flocks at slaughter, Southern island of Japan, June 1998 to December 2003
970
SHAHADA ET AL.
Table 2. Status of Salmonella infection in relation to the frequency of sampling of the holdings Sampling frequency Low (once) (2 to 4) Medium (5 to 9) High (10 to 14) Total
Status of infection1 Clean
Intermittent
Persistent
Total holdings
5 9 0 0 14
— 11 5 3 19
92 9 7 5 30
14 29 12 8 63
1 Holdings were classified as clean, intermittent, and persistent when Salmonella detection in flocks was negative, fluctuating negative and positive, and throughout (100%) positive, respectively. 2 Flocks sampled once and tested positive.
lence were found to be similar (approximately 71%), and seasonality component was not observed.
Detection of Salmonella in relation to the frequency of sampling of the holdings was determined as summarized in Table 2. The status varied from Salmonella-negative holdings (n = 9), holdings that showed intermittent infections with alternating negative and positive flocks (n = 19; Table 3), and those with persistent infections in which all flocks tested positive (n = 21). Fourteen holdings (negative, n = 5 and positive, n = 9) were sampled only once throughout the study period. A positive relationship between frequency of sampling and status of infection of the holdings was observed. Frequently sampled holdings were found to be contaminated intermittently or persistently. In contrary, all apparently clean holdings were found in the category of holdings less frequently sampled.
PFGE Analysis of Salmonella Infantis Because Salmonella Infantis was the most frequent isolate, 36 randomly chosen representatives were analyzed
DISCUSSION To date there is limited information pertaining to the actual prevalence of Salmonella in broiler flocks from production to processing. This report presents important information obtained through routine Salmonella surveillance at the poultry processing plant. We established temporal distribution of the infection and determined the genetic fingerprints of the predominant serovar in broilers. Seventy-one percent of the flocks were positive, and salmonellae were isolated from 14% of cecal samples surveyed. Previous studies on this topic have shown comparable Salmonella isolation rate within-flocks elsewhere (Limawongpranee et al., 1999; Evans and Wegener, 2003). Mandatory Salmonella surveillance has been implemented in some countries (Gradel et al., 2002), and as from the start of the surveillance programs in those countries cecal samples from 16 chickens per flock were tested. Assuming a test sensitivity of 1.00, this sampling procedure could with 90% confidence detect Salmonella-positive flocks with a prevalence of 20% (Bisgaard, 1992; Skov et al., 1999). We detected high prevalence of Salmonellapositive flocks by employing similar sampling procedure and therefore its validation. In the present survey, the prevalence of Salmonellapositive flocks did not vary significantly throughout the study period and seasonality component was not observed. A small proportion of flocks were raised in Salmo-
Figure 1. Salmonella Infantis pulsed field gel electrophoresis (PFGE) patterns represented as (A) XbaI common genotype (lanes 1-12) and (B) BlnI genotype patterns I (lanes 1, 4, 6, 7, 8 and 12), Ia (lane 3), Ib (lane 5, 10 and 11), II (lane 2), and III (lane 9). M, molecular (lambda) marker. Lanes 1 to 12 represent isolates SI 25 to 36.
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
Persistence of Salmonella in Farms
by PFGE to determine their genetic relatedness. Digestion by XbaI yielded 1 common genotype in all isolates (Figure 1A), whereas BlnI produced 3 (I-III) genotypes (Figure 1B). Most of the BlnI DNA fingerprints (87.5%; 21/24) were grouped into genotype I, 2 banding patterns were classified as genotype II, and 1 pattern formed genotype III (Table 4). Differences in band number of small DNA fragments (Figure 1B) led to further division of genotype I into subtypes Ia (n = 3) and Ib (n = 1).
PREVALENCE AND GENETIC FINGERPRINTS OF SALMONELLA Table 3. Status of Salmonella in sampled flocks from 8 different holdings (designated A to H) with fluctuating negative and positive flocks Sampling frequency1 Farm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
A B C D E F G H
+ + − − + + + +
− − − + + + + +
− − + − + + + −
− − + + + + + +
+ + + + + + + +
− − − − + + + +
− − + + + − + +
− − + − + − − −
− − − + − − −
+ + − −
− − + +
+ + +
+ + −
+ +
1
The sampling frequency from A to D holdings was high (7 to 9) while from E to H holdings was medium (11 to 14). Fluctuation was not significant at the low sampling frequency class and hence not included.
Table 4. Pulsed-field gel electrophoresis patterns of broiler Salmonella Infantis, 1998 to 2003 Genotype patterns for Strain designation SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI
01-SI 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Isolation mo/yr
XbaI
Bln1
1998 to 1999 Jan. 2000 Mar. 2000 May 2000 July 2000 Sep. 2000 Nov. 2000 Feb. 2001 Apr. 2001 June 2001 Aug. 2001 Oct. 2001 Dec. 2001 Jan. 2002 Mar. 2002 May 2002 July 2002 Sep. 2002 Nov. 2002 Feb. 2003 Apr. 2003 June 2003 Aug. 2003 Oct. 2003 Dec. 2003
I I I I I I I I I I I I I I I I I I I I I I I I I
Not done I Ia I I I I Ia I I II I I I II Ia I I I I I III I Ib I
pending on the status, a control action plan tailored for individual farms is effected. Farms characterized by intermittent infections may have implemented routine preventive measures as planned. However, intermittent infection trends suggest that such control measures carried out have failed to prevent reinfection with Salmonella. Again, farms characterized by persistent infections are likely to have persistent bacterial infectious sources in the environment. It has been observed that elimination of Salmonella from broiler flocks often is difficult due to many sources of contamination within the farm environment (Davies and Wray, 1996). Analysis of DNA macrorestriction fingerprinting patterns generated by PFGE indicated Salmonella Infantis isolates to share the same genotype, and therefore, likelihood of a common source of infection. Moreover, related studies (Hamada et al., 2003b; Kudaka et al., 2006; Asai et al., 2007) described Salmonella Infantis as the most frequent isolate in broilers, and their genotypes showed a close resemblance. Despite the chromosome profile similarity observed with XbaI, we revealed changes consistent with a single genetic event when BlnI was used. Such random genetic events normally occur during the course of time and lead to alteration of PFGE pattern. We considered the minor changes to be nonsignificant, and because the isolates colonized and persisted at the broiler industry since the late 1990s, we inferred them to be clonally and epidemiologically related. The findings provide evidence that support presumptions that the isolates may be derived from a common source of infection and similar time of dissemination. Besides, high levels of colonization ability and genetic stability were observed. The attributes of most of these events remain largely unknown to date and hence call for further investigations. Poultry is widely implicated as an important cause of Salmonella infections in humans and salmonellae found in chickens are often isolated from humans (Tauxe, 1991; Morris, 1996; Ekperigin and Nagaraja, 1998). In the present study the most frequent serotype recovered in chicken was not the most common serovar isolated from cases of human salmonellosis. The most common isolate from chicken since 1997 has been Salmonella Infantis, whereas in humans the most common isolate was Salmonella Enteritidis (Asai et al., 2007). According to the recent Japanese Infectious Agents Surveillance report, Salmonella Infantis has been the second most common human isolate (http:// idsc.nih.go.jp/iasr/index.html Accessed Nov. 2007). It is noteworthy the fact that Salmonella Infantis is highly prevalent in poultry and humans suggests a relationship, regardless of the prevalences of other serotypes. Salmonella Enteritidis infection in humans in Japan is most often acquired from the consumption of raw or undercooked eggs. In conclusion, this study provided insights that the isolates might have been derived from a common source and the same time of dissemination of infection. Feedback of such epidemiological information to producers would enable the formulation of effective control action plan
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
nella-free holdings (22%), whereas the majority of flocks originated from intermittently or persistently infected holdings. Farms with medium to high sampling frequency were generally characterized by the contamination manifested as intermittent, persistent, or both, Salmonella infections. It is likely that the observed Salmonellafree holdings are provisionally clean and that increased flock-turnover may lead to contamination because all infected farms revealed medium to high number of flock output. It is noteworthy that where Salmonella testing is not a mandate, monitoring and surveillance systems mainly operate through an abattoir-farm scheme whereby abattoirs return the information about the status of the infection to producers and that they test their flocks voluntarily. Thus, in abattoir-farm schemes, producers mainly rely on slaughterhouse feedback information, and de-
971
972
SHAHADA ET AL.
tailored for individual holdings with special emphasis on biosecurity, hygiene, and pest control.
REFERENCES
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
Asai, T., K. Ishihara, K. Harada, A. Kojima, Y. Tamura, S. Sato, and T. Takahashi. 2007. Long-term prevalence of antimicrobial-resistant Salmonella enterica subspecies enterica Serovar Infantis in the broiler chicken industry in Japan. Microbiol. Immunol. 51:111–115. Bisgaard, M. 1992. A voluntary Salmonella control programme for the broiler industry, implemented by the Danish Poultry Council. Int. J. Food Microbiol. 15:219–224. Davies, R. H., and C. Wray. 1996. Persistence of Salmonella enteritidis in poultry units and poultry food. Br. Poult. Sci. 37:589–596. Ekperigin, H. E., and K. V. Nagaraja. 1998. Microbial food borne pathogens. Salmonella. Vet. Clin. North Am. Food Anim. Pract. 14:17–29. Evans, M. C., and H. C. Wegener. 2003. Antimicrobial growth promoters and Salmonella spp., Campylobacter spp. in poultry and swine, Denmark. Emerg. Infect. Dis. 9:489–492. Gradel, K. O., J. Andersen, and M. Madsen. 2002. Comparisons of sampling procedures and time of sampling for the detection of Salmonella in Danish infected chicken flocks raised in floor systems. Acta Vet. Scand. 43:21–30. Hamada, K., K. Oshima, and H. Tsuji. 2003a. Drug resistance genes encoded in integrons and in extra-integrons: Their distribution and lateral transfer among pathogenic enterobacteriaceae including enterohemorrhagic Escherichia coli and Salmonella enterica serovars Typhimurium and Infantis. Jpn. J. Infect. Dis. 56:123–126. Hamada, K., K. Oshima, and H. Tsuji. 2003b. Chromosomal transferable multidrug resistance genes of Salmonella enterica serovar Infantis. Jpn. J. Infect. Dis. 56:216–218. Humphrey, T. J., E. J. Threlfall, and J. G. Cruickshank. 1998. Salmonellosis. Pages 191–206 in Zoonoses: Biology, Clinical
Practice, and Public Health Control. S. R. Palmer, E. J. L. Soulsby, and D. I. H. Simpson, ed. Oxford University Press, New York. Khakhria, R., D. Woodward, W. M. Johnson, and C. Poppe. 1997. Salmonella isolated from humans, animals and other sources in Canada, 1983–1992. Epidemiol. Infect. 119:15–23. Kudaka, J., K. Itokazu, K. Taira, A. Iwai, M. Kondo, T. Susa, and M. Iwanaga. 2006. Characterization of Salmonella isolated in Okinawa, Japan. Jpn. J. Infect. Dis. 59:15–19. Limawongpranee, S., H. Hayashidani, A. T. Okatani, K. Ono, C. Hirota, K. Kaneko, and M. Ogawa. 1999. Prevalence and persistence of Salmonella in broiler chicken flocks. J. Vet. Med. Sci. 61:255–259. Morris, J. G., Jr. 1996. Current trends in human diseases associated with foods of animal origin. J. Am. Vet. Med. Assoc. 209:2045–2047. Olsen, S. J., R. Bishop, and F. W. Brenner. 2001. The changing epidemiology of Salmonella: Trends in serotypes isolated from humans in the United States, 1987–1997. J. Infect. Dis. 183:753–761. Rambach, A. 1990. New plate medium for facilitated differentiation of Salmonella spp. from Proteus spp. and other enteric bacteria. Appl. Environ. Microbiol. 56:301–303. Ribot, E. M., M. A. Fair, R. Gautom, D. N. Cameron, S. B. Hunter, B. Swaminathan, and T. J. Barrett. 2006. Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli O157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog. Dis. 3:59–67. Shahada, F., A. Amamoto, T. Chuma, A. Shirai, and K. Okamoto. 2007. Antimicrobial susceptibility phenotypes, resistance determinants and DNA fingerprints of Salmonella enterica serotype Typhimurium isolated from bovine in Japan. Int. J. Antimicrob. Agents 30:150–156. Skov, M. N., B. Carstensen, N. Tornøe, and M. Madsen. 1999. Evaluation of sampling methods for the detection of Salmonella in broiler flocks. J. Appl. Microbiol. 86:695–700. Tauxe, R. 1991. Salmonella: A postmodern pathogen. J. Food Prot. 54:563–568.
Key words: Salmonella, broiler flock, temporal distribution, serotype, genotype 2008 Poultry Science 87:968–972 doi:10.3382/ps.2007-00455
Meat processing plants are potential hubs in the detection of infectious disease agents including Salmonella. Determining the Salmonella status of broiler flocks during slaughter is necessary to direct interventions at high prevalence holdings. Establishing the status of Salmonella isolates has been useful as a way of following trends over time, and also an important way in identifying common isolates that cluster over time or space (Olsen et al., 2001). A temporal cluster is a group of isolates of a particular serovar that aggregate together in time. Thus, identifying clusters can provide information that may be used to establish possible determinants of the disease and methods that can be used for control and prevention (Tauxe, 1991). Currently, there is limited information pertaining to the contamination of national and regional poultry facilities by Salmonella. Being part of the veterinary infectious agents monitoring and surveillance system in Japan, we report for the first time the incidence, temporal distribution, and genetic fingerprints of Salmonella in broilers in the Southern prefecture of Japan’s main island. A systematic abattoir longitudinal study was conducted to estab-
INTRODUCTION Serotypes of non-Typhi Salmonella enterica (hereafter referred to as Salmonella) are important pathogenic organisms in humans and other animals (Khakhria et al., 1997). Subclinical infection is not uncommon, and many animals, particularly poultry, become asymptomatic carriers, intermittently shedding the bacteria into the environment for variable periods of time. The latter situation is particularly important in the transmission and maintenance of Salmonella (Humphrey et al., 1998). Nontyphoid human infections are most often acquired from contaminated foods of animal origin. Salmonella Infantis and Salmonella Typhimurium have been major causes of salmonellosis in humans in Japan until 1989 when Salmonella Enteritidis suddenly emerged and continued prevailing to date (Hamada et al., 2003a; Asai et al., 2007).
©2008 Poultry Science Association Inc. Received November 7, 2007. Accepted January 29, 2008. 1 Corresponding author: [email protected]
968
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
holdings (n = 9), positive-flocks from persistently infected holdings (n = 21), and holdings (n = 19) that showed fluctuations with alternating negative and positive flocks for variable time periods. Fourteen holdings (negative, n = 5 and positive, n = 9) were sampled once throughout the study period. Seasonality component was not observed, and salmonellae were found colonizing broiler ceca in warm and cold months. Predominant serovar was Salmonella Infantis (93.3%; n = 525). Macrorestriction fingerprints of Salmonella Infantis using XbaI presumed the isolates to be derived from a common parent. Enhanced discrimination by BlnI digestion produced 3 banding patterns that were closely related genetically and hence epidemiologically related. Such epidemiological information may enable producers to formulate effective control action plan tailored for individual holdings with special emphasis on biosecurity, hygiene, and pest control.
ABSTRACT During the 1998 to 2003 period, cecal contents of 4,024 broiler chickens from 252 flocks raised in 63 holdings were examined for Salmonella. The aims were to establish the actual status of the infection, its temporal distribution, prevalent serotype, and common genotype among broiler flocks brought at the slaughterhouse. Collected samples were preenriched in Hajna tetrathionate broth, and after 24 h of incubation, 10 L of the broth was streaked on selective Rambach agar plate. Suspected scarlet color colonies of Salmonella were cloned on nutrient agar, confirmed through biochemical tests and serotyped using O and H antigens. Pulsed field gel electrophoresis technique generated DNA fragments banding patterns and established their clonal relatedness. Salmonella was isolated from 563 (14%) samples in 179 (71%) flocks. The flock situation varied from Salmonella-negative
969
PREVALENCE AND GENETIC FINGERPRINTS OF SALMONELLA
lish the percentage of positive flocks and identify the temporal trends and seasonality of Salmonella infections in broiler chickens.
MATERIALS AND METHODS Sample Collection and Preparation
Bacterial Isolation Approximately 1 g of cecal contents was aseptically mixed with 5 mL of sterilized distilled water and homogenized by vortexing. Then, 1 mL of the suspension was preenriched in 5 mL of Hajna tetrathionate broth (Eiken Chemical Co. Ltd., Tokyo, Japan) and incubated in a water bath at 42°C. After 24 h of incubation, a loopful from each of the enriched broth was streaked onto plates of selective Rambach (Rambach, 1990) agar, and incubated at 37°C for 24 h.
Bacteria Identification Suspected scarlet color colonies were selected from each plate and cloned on nutrient agar slants. Salmonella were confirmed using Gram staining and biochemical tests including fermentation of glucose, lactose and sucrose, hydrogen sulfide production, citrate, lysine decarboxylation, methyl red, and indole tests. Gram-negative rods with typical biochemical profiles of Salmonella were serotyped for somatic antigen by slide agglutination test using Salmonella group O antisera (polyvalent O, O4, O7, and O9). Salmonella isolates were further serotyped for flagellar (H) antigens by the tube agglutination test according to the Kaufmann-White scheme.
Pulsed Field Gel Electrophoresis A random sample (n = 36) of predominant O7 serogroup was drawn and DNA analyzed by pulsed field gel electrophoresis (PFGE). From June to December 1998, 1 sample was selected each month. From 1999 through 2003, 1 sample was drawn after every 2 mo. To trace the genetic diversity, samples were selected from different farms. Plugs preparation, XbaI and BlnI digestion were
Year
n samples
n flocks
n positive samples (%)
1998 1999 2000 2001 2002 2003 Total
381 732 768 704 704 735 4,024
24 46 48 44 44 46 252
57 (15) 116 (15.8) 115 (15) 88 (12.5) 76 (10.8) 111 (15.1) 563 (14)
carried out according to the procedure described by Ribot et al. (2006). Electrophoresis conditions, staining of gels, and digital images acquisition were performed as previously described by Shahada et al. (2007).
Data Analysis Statistical analysis was performed in an Excel spreadsheet. The percentage of positive flocks was calculated, and differences in mean prevalence were evaluated by using Microsoft Excel Analyze-It statistical tests for the χ2 and ANOVA. The probability level of 0.05 was used throughout. Analysis and interpretation of PFGE TIFF images was carried out as previous explained (Shahada et al., 2007).
RESULTS Prevalence of Salmonella Seventy-one percent of broiler flocks tested Salmonellapositive, and Salmonella was detected in 563 (14 %) out of 4,024 samples (Table 1). Out of 563 positive samples, 525 (93.3%) were serogroup O7, 6 (1.1%) serogroup O9, 2 (0.4%) serogroup O4, 10 (1.8%) polyvalent O, and 20 (3.55%) were not typed. Determination of H antigens in serogroup O7 revealed phase I (r) antigens produced by motile Salmonella Infantis and, hence, formed the predominant serotype. Serogroups O4 and O9 comprise S. Typhimurium and Salmonella Enteritidis serotypes, respectively. Of 252 flocks examined, 171 were positive with a single Salmonella serotype, whereas 8 revealed mixed serotypes. A comparison of means showed there was no statistical difference in the mean prevalence between the flocks.
Seasonality Component Salmonella prevalence trends did not vary when the 12mo time periods (January to December) were compared. A similar trend was also depicted by the cumulative 5yr period from January 1999 through December 2003. Minimum and maximum annual flock prevalence ranged between 56.8 and 79.5% during 2002 and 2001, respectively, whereas the cumulative monthly prevalence ranged between 56.5 and 82.6% in July and October, respectively. As a result, monthly and annually flock preva-
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
Samples were selected from 2 flocks on each sampling day. Sampled flocks were selected at random among 5 to 6 flocks slaughtered daily at the plant. Sample collection was carried out systematically at 14-d intervals throughout the study period. A total of 4,024 samples derived from 252 flocks were collected following 127 visits to the plant. Seventy-four of these flocks from 43 holdings were sampled 1 to 4 times; 84 flocks from 12 holdings were sampled 5 to 9 times; and 94 flocks originating from 8 holdings were tested 10 to 14 times. All samples were taken using sterile techniques, placed in sterile sampling plastic bags, and chilled by ice blocks during transportation. The samples were cultured on the same day upon arrival at the laboratory.
Table 1. Status of Salmonella in broiler flocks at slaughter, Southern island of Japan, June 1998 to December 2003
970
SHAHADA ET AL.
Table 2. Status of Salmonella infection in relation to the frequency of sampling of the holdings Sampling frequency Low (once) (2 to 4) Medium (5 to 9) High (10 to 14) Total
Status of infection1 Clean
Intermittent
Persistent
Total holdings
5 9 0 0 14
— 11 5 3 19
92 9 7 5 30
14 29 12 8 63
1 Holdings were classified as clean, intermittent, and persistent when Salmonella detection in flocks was negative, fluctuating negative and positive, and throughout (100%) positive, respectively. 2 Flocks sampled once and tested positive.
lence were found to be similar (approximately 71%), and seasonality component was not observed.
Detection of Salmonella in relation to the frequency of sampling of the holdings was determined as summarized in Table 2. The status varied from Salmonella-negative holdings (n = 9), holdings that showed intermittent infections with alternating negative and positive flocks (n = 19; Table 3), and those with persistent infections in which all flocks tested positive (n = 21). Fourteen holdings (negative, n = 5 and positive, n = 9) were sampled only once throughout the study period. A positive relationship between frequency of sampling and status of infection of the holdings was observed. Frequently sampled holdings were found to be contaminated intermittently or persistently. In contrary, all apparently clean holdings were found in the category of holdings less frequently sampled.
PFGE Analysis of Salmonella Infantis Because Salmonella Infantis was the most frequent isolate, 36 randomly chosen representatives were analyzed
DISCUSSION To date there is limited information pertaining to the actual prevalence of Salmonella in broiler flocks from production to processing. This report presents important information obtained through routine Salmonella surveillance at the poultry processing plant. We established temporal distribution of the infection and determined the genetic fingerprints of the predominant serovar in broilers. Seventy-one percent of the flocks were positive, and salmonellae were isolated from 14% of cecal samples surveyed. Previous studies on this topic have shown comparable Salmonella isolation rate within-flocks elsewhere (Limawongpranee et al., 1999; Evans and Wegener, 2003). Mandatory Salmonella surveillance has been implemented in some countries (Gradel et al., 2002), and as from the start of the surveillance programs in those countries cecal samples from 16 chickens per flock were tested. Assuming a test sensitivity of 1.00, this sampling procedure could with 90% confidence detect Salmonella-positive flocks with a prevalence of 20% (Bisgaard, 1992; Skov et al., 1999). We detected high prevalence of Salmonellapositive flocks by employing similar sampling procedure and therefore its validation. In the present survey, the prevalence of Salmonellapositive flocks did not vary significantly throughout the study period and seasonality component was not observed. A small proportion of flocks were raised in Salmo-
Figure 1. Salmonella Infantis pulsed field gel electrophoresis (PFGE) patterns represented as (A) XbaI common genotype (lanes 1-12) and (B) BlnI genotype patterns I (lanes 1, 4, 6, 7, 8 and 12), Ia (lane 3), Ib (lane 5, 10 and 11), II (lane 2), and III (lane 9). M, molecular (lambda) marker. Lanes 1 to 12 represent isolates SI 25 to 36.
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
Persistence of Salmonella in Farms
by PFGE to determine their genetic relatedness. Digestion by XbaI yielded 1 common genotype in all isolates (Figure 1A), whereas BlnI produced 3 (I-III) genotypes (Figure 1B). Most of the BlnI DNA fingerprints (87.5%; 21/24) were grouped into genotype I, 2 banding patterns were classified as genotype II, and 1 pattern formed genotype III (Table 4). Differences in band number of small DNA fragments (Figure 1B) led to further division of genotype I into subtypes Ia (n = 3) and Ib (n = 1).
PREVALENCE AND GENETIC FINGERPRINTS OF SALMONELLA Table 3. Status of Salmonella in sampled flocks from 8 different holdings (designated A to H) with fluctuating negative and positive flocks Sampling frequency1 Farm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
A B C D E F G H
+ + − − + + + +
− − − + + + + +
− − + − + + + −
− − + + + + + +
+ + + + + + + +
− − − − + + + +
− − + + + − + +
− − + − + − − −
− − − + − − −
+ + − −
− − + +
+ + +
+ + −
+ +
1
The sampling frequency from A to D holdings was high (7 to 9) while from E to H holdings was medium (11 to 14). Fluctuation was not significant at the low sampling frequency class and hence not included.
Table 4. Pulsed-field gel electrophoresis patterns of broiler Salmonella Infantis, 1998 to 2003 Genotype patterns for Strain designation SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI SI
01-SI 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Isolation mo/yr
XbaI
Bln1
1998 to 1999 Jan. 2000 Mar. 2000 May 2000 July 2000 Sep. 2000 Nov. 2000 Feb. 2001 Apr. 2001 June 2001 Aug. 2001 Oct. 2001 Dec. 2001 Jan. 2002 Mar. 2002 May 2002 July 2002 Sep. 2002 Nov. 2002 Feb. 2003 Apr. 2003 June 2003 Aug. 2003 Oct. 2003 Dec. 2003
I I I I I I I I I I I I I I I I I I I I I I I I I
Not done I Ia I I I I Ia I I II I I I II Ia I I I I I III I Ib I
pending on the status, a control action plan tailored for individual farms is effected. Farms characterized by intermittent infections may have implemented routine preventive measures as planned. However, intermittent infection trends suggest that such control measures carried out have failed to prevent reinfection with Salmonella. Again, farms characterized by persistent infections are likely to have persistent bacterial infectious sources in the environment. It has been observed that elimination of Salmonella from broiler flocks often is difficult due to many sources of contamination within the farm environment (Davies and Wray, 1996). Analysis of DNA macrorestriction fingerprinting patterns generated by PFGE indicated Salmonella Infantis isolates to share the same genotype, and therefore, likelihood of a common source of infection. Moreover, related studies (Hamada et al., 2003b; Kudaka et al., 2006; Asai et al., 2007) described Salmonella Infantis as the most frequent isolate in broilers, and their genotypes showed a close resemblance. Despite the chromosome profile similarity observed with XbaI, we revealed changes consistent with a single genetic event when BlnI was used. Such random genetic events normally occur during the course of time and lead to alteration of PFGE pattern. We considered the minor changes to be nonsignificant, and because the isolates colonized and persisted at the broiler industry since the late 1990s, we inferred them to be clonally and epidemiologically related. The findings provide evidence that support presumptions that the isolates may be derived from a common source of infection and similar time of dissemination. Besides, high levels of colonization ability and genetic stability were observed. The attributes of most of these events remain largely unknown to date and hence call for further investigations. Poultry is widely implicated as an important cause of Salmonella infections in humans and salmonellae found in chickens are often isolated from humans (Tauxe, 1991; Morris, 1996; Ekperigin and Nagaraja, 1998). In the present study the most frequent serotype recovered in chicken was not the most common serovar isolated from cases of human salmonellosis. The most common isolate from chicken since 1997 has been Salmonella Infantis, whereas in humans the most common isolate was Salmonella Enteritidis (Asai et al., 2007). According to the recent Japanese Infectious Agents Surveillance report, Salmonella Infantis has been the second most common human isolate (http:// idsc.nih.go.jp/iasr/index.html Accessed Nov. 2007). It is noteworthy the fact that Salmonella Infantis is highly prevalent in poultry and humans suggests a relationship, regardless of the prevalences of other serotypes. Salmonella Enteritidis infection in humans in Japan is most often acquired from the consumption of raw or undercooked eggs. In conclusion, this study provided insights that the isolates might have been derived from a common source and the same time of dissemination of infection. Feedback of such epidemiological information to producers would enable the formulation of effective control action plan
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
nella-free holdings (22%), whereas the majority of flocks originated from intermittently or persistently infected holdings. Farms with medium to high sampling frequency were generally characterized by the contamination manifested as intermittent, persistent, or both, Salmonella infections. It is likely that the observed Salmonellafree holdings are provisionally clean and that increased flock-turnover may lead to contamination because all infected farms revealed medium to high number of flock output. It is noteworthy that where Salmonella testing is not a mandate, monitoring and surveillance systems mainly operate through an abattoir-farm scheme whereby abattoirs return the information about the status of the infection to producers and that they test their flocks voluntarily. Thus, in abattoir-farm schemes, producers mainly rely on slaughterhouse feedback information, and de-
971
972
SHAHADA ET AL.
tailored for individual holdings with special emphasis on biosecurity, hygiene, and pest control.
REFERENCES
Downloaded from http://ps.oxfordjournals.org/ at D H Hill Library - Acquis S on March 29, 2015
Asai, T., K. Ishihara, K. Harada, A. Kojima, Y. Tamura, S. Sato, and T. Takahashi. 2007. Long-term prevalence of antimicrobial-resistant Salmonella enterica subspecies enterica Serovar Infantis in the broiler chicken industry in Japan. Microbiol. Immunol. 51:111–115. Bisgaard, M. 1992. A voluntary Salmonella control programme for the broiler industry, implemented by the Danish Poultry Council. Int. J. Food Microbiol. 15:219–224. Davies, R. H., and C. Wray. 1996. Persistence of Salmonella enteritidis in poultry units and poultry food. Br. Poult. Sci. 37:589–596. Ekperigin, H. E., and K. V. Nagaraja. 1998. Microbial food borne pathogens. Salmonella. Vet. Clin. North Am. Food Anim. Pract. 14:17–29. Evans, M. C., and H. C. Wegener. 2003. Antimicrobial growth promoters and Salmonella spp., Campylobacter spp. in poultry and swine, Denmark. Emerg. Infect. Dis. 9:489–492. Gradel, K. O., J. Andersen, and M. Madsen. 2002. Comparisons of sampling procedures and time of sampling for the detection of Salmonella in Danish infected chicken flocks raised in floor systems. Acta Vet. Scand. 43:21–30. Hamada, K., K. Oshima, and H. Tsuji. 2003a. Drug resistance genes encoded in integrons and in extra-integrons: Their distribution and lateral transfer among pathogenic enterobacteriaceae including enterohemorrhagic Escherichia coli and Salmonella enterica serovars Typhimurium and Infantis. Jpn. J. Infect. Dis. 56:123–126. Hamada, K., K. Oshima, and H. Tsuji. 2003b. Chromosomal transferable multidrug resistance genes of Salmonella enterica serovar Infantis. Jpn. J. Infect. Dis. 56:216–218. Humphrey, T. J., E. J. Threlfall, and J. G. Cruickshank. 1998. Salmonellosis. Pages 191–206 in Zoonoses: Biology, Clinical
Practice, and Public Health Control. S. R. Palmer, E. J. L. Soulsby, and D. I. H. Simpson, ed. Oxford University Press, New York. Khakhria, R., D. Woodward, W. M. Johnson, and C. Poppe. 1997. Salmonella isolated from humans, animals and other sources in Canada, 1983–1992. Epidemiol. Infect. 119:15–23. Kudaka, J., K. Itokazu, K. Taira, A. Iwai, M. Kondo, T. Susa, and M. Iwanaga. 2006. Characterization of Salmonella isolated in Okinawa, Japan. Jpn. J. Infect. Dis. 59:15–19. Limawongpranee, S., H. Hayashidani, A. T. Okatani, K. Ono, C. Hirota, K. Kaneko, and M. Ogawa. 1999. Prevalence and persistence of Salmonella in broiler chicken flocks. J. Vet. Med. Sci. 61:255–259. Morris, J. G., Jr. 1996. Current trends in human diseases associated with foods of animal origin. J. Am. Vet. Med. Assoc. 209:2045–2047. Olsen, S. J., R. Bishop, and F. W. Brenner. 2001. The changing epidemiology of Salmonella: Trends in serotypes isolated from humans in the United States, 1987–1997. J. Infect. Dis. 183:753–761. Rambach, A. 1990. New plate medium for facilitated differentiation of Salmonella spp. from Proteus spp. and other enteric bacteria. Appl. Environ. Microbiol. 56:301–303. Ribot, E. M., M. A. Fair, R. Gautom, D. N. Cameron, S. B. Hunter, B. Swaminathan, and T. J. Barrett. 2006. Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli O157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog. Dis. 3:59–67. Shahada, F., A. Amamoto, T. Chuma, A. Shirai, and K. Okamoto. 2007. Antimicrobial susceptibility phenotypes, resistance determinants and DNA fingerprints of Salmonella enterica serotype Typhimurium isolated from bovine in Japan. Int. J. Antimicrob. Agents 30:150–156. Skov, M. N., B. Carstensen, N. Tornøe, and M. Madsen. 1999. Evaluation of sampling methods for the detection of Salmonella in broiler flocks. J. Appl. Microbiol. 86:695–700. Tauxe, R. 1991. Salmonella: A postmodern pathogen. J. Food Prot. 54:563–568.