 |
Rural Industries Research & Development Corporation |
|| Home || Search || Contact || Publications Eshop || Privacy Statement ||
Rural Industries Research & Development Corporation
Summary of full report
Molecular Basis of Benign Colonisation
of Salmonella Sofia in Chickens
August 2001
RIRDC Publication No 01/106 RIRDC Project No IMVS-1A
Executive Summary
Background to the project
Contamination of poultry products by bacteria can lead to public health problems, and adverse media coverage of such incidents has heightened consumer awareness of food safety. These reports can unnecessarily affect retail consumption of implicated food/food products. In view of these challenges that face the industry, the development of rapid typing methods for Salmonella in conjunction with classical methods was initiated in 1985 with CMRDC funding. Molecular methods can often offer greater discriminating power between human and poultry isolates and can be used for epidemiological studies of disease transmission.
Salmonella Sofia first came to the attention of the Australian Salmonella Reference Centre in 1979 as a new isolate from chickens. Despite the widespread colonisation of chickens by S. Sofia it is not represented in the list of serovars isolated from humans, which indicates it may be of low virulence to humans. Currently it represents 58% (1999 figures) of all Salmonella in chickens of the approximately 2,500 - 3000 isolates submitted annually to the IMVS for typing from the poultry industry. S. Sofia has now been the predominant Salmonella from chickens in Australia for nearly two decades.
S. Sofia is ubiquitous amongst Australian chicken flocks and few serious Salmonella food poisoning outbreaks attributed to chicken meat have occurred since S. Sofia has become the predominant isolate from chickens. These observations have lead to the idea that S. Sofia may be acting as natural competitive exclusion (CE) agent. CE agents are either mixtures or pure cultures of natural benign gut bacteria that through their colonisation of the gut exclude pathogenic bacteria. The commercial preparation Broilact. that is a mixture of many bacterial species is a well-known CE agent.
This project aims to determine whether S. Sofia can exclude virulent serovars of Salmonella bacteria, as well as monitoring the distribution of Salmonella serovars in Australian chickens by conventional phage typing, serotyping and DNA fingerprinting (molecular typing).
Methodology
Conventional and Molecular typing of Salmonella Conventional and molecular typing services using pulsed field gel electrophoresis (PFGE) are already well established within the laboratory, relevant strains from industry sources were processed on an on going basis as they were received.
Colonisation trials in chickens with natural S. Sofia variants
Previous work on project IMVS6CM has shown that S. Sofia lacks a complete set of virulence genes and that the extent of the genetic lesions is highly variable. Therefore, S. Sofia strains were chosen based upon their degree of genetic incompleteness to use in colonisation studies of chickens at the Victorian Institute of Animal Science. S. Sofia strains that are able to colonise chickens for extended periods (up to 36 days) were then chosen further studies. These studies which examined the ability of S. Sofia pre-colonised birds to exclude S. Typhimurium from chickens are described below.
Challenge of S. Sofia strains selected for the ability to colonise chickens for long periods by virulent serovars
Salmonella-free birds were first inoculated with S. Sofia chosen from the previous study, and measurement of colonisation status with respect to S. Sofia was carried out over a number of days.
These birds were then be challenged with S. Typhimurium, by placing birds that were pre-colonised with S. Typhimurium in the same isolator as the birds pre-colonised with S. Sofia. The health of the birds was monitored closely and the extent of colonisation of the S. Typhimurium and S. Sofia determined in both challenge and pre-colonised birds.
PCR based survey in S. Sofia of known Salmonella colonisation factors
The mechanism of colonisation of chickens by S. Sofia is most probably determined by a fimbrial structure (colonisation factor) on the bacterial cell surface. The DNA sequences of the genes encoding many types of fimbriae in Salmonella are available. The existence of a number of these genes was determined in S. Sofia by PCR and Southern hybridisation.
Knockout mutagenesis of colonisation factor genes and construction of isogenic mutants in S. Sofia.
Fimbrial genes for SEF 17 that appeared to be present in all S. Sofia isolates were chosen. These genes were then cloned using PCR. The DNA sequences of the cloned genes was then determined by automated DNA sequencing and compared with similar genes in other Salmonella serovars from computer databases. The cloned gene that encoded the major fimbrial protein of SEF17 was inactivated by insertion of foreign DNA. This mutation should prevent expression of the fimbriae on the surface of S. Sofia. The mutated gene was then introduced and recombined into the chromosome of S. Sofia using a suicide vector.
Trial of S. Sofia colonisation factor mutants for colonisation of chickens
This task was not carried out due to time restraints,but all prerequisites were met.
Results
Conventional and Molecular typing - Overview 1997 - 2000
These figures represent an average of 3,332 specimens received from the industry over a five-year period. The variability in the number of specimens received may be a result of basic typing for S.
Sofia occurring within industry laboratories. The decline in S. Sofia numbers observed in recent years has reversed or stabilised. In 1999 S. Sofia represented 57.8% of Salmonella isolated from chickens. It is still rarely isolated from humans. The predominant S. Typhimurium phage types isolated from chickens and humans remain 9, 64 and 135.
Several outbreaks caused by Salmonella have occurred during the grant period and have had high media profiles. None of these incidents involved poultry products. A concern to public health authorities is the spread of S. Virchow across Australia. This serovar has been a predominant serovar in Queensland since the mid -1970s and remained localised to that region. In recent years cases have occurred with increasing frequency in the southern Australian states. Phage types 8 and 34 are found frequently in the human population in Australia. A reservoir for the spread has not been identified.
The Australian Salmonella Reference Centre is now phage typing all S. Virchow isolates, using the Division of Enteric Pathogens, Colindale, U.K., typing scheme. Further analysis of these strains using RFLP techniques has revealed that there are at least seven different RFLP types by PFGE among phage type 8 isolates and that phage type 34 is related to phage type 8.
Colonisation trials in chickens with natural S. Sofia variants
These experiments identified five S. Sofia strains that were able to colonise chickens for extended periods, these strains were SSWT, MH7, MH76, MH89 and MH107. These strains were able to colonise chickens for up to 36 days. There appeared to be no correlation with the genetic characteristics identified in this and previous reports and the ability to colonise chickens. The above strains were then used in challenge studies using S. Typhimurium.
Challenge of S. Sofia strains selected for the ability to colonise chickens for long periods by virulent serovars
All S Sofia groups became infected with S. Typhimurium after contact with S. Typhimurium pre- infected birds. Strain MH76 infected birds showed infection with both S. Sofia and S. Typhimurium from day 12 onward to the end of the experiment at day 36 in two independent experiments. Strain MH76 also persists in the presence of S. Typhimurium and is able to co-colonise birds that were initially colonised with S. Typhimurium. It is also notable the other strains SSWT, MH107, MH89 and MH7 were not isolated for any significant period after contact with S. Typhimurium infected birds. In summary S. Sofia is not acting as a competitive exclusion (CE) agent, but can co-colonise with S. Typhimurium. The mechanism behind the persistence of MH76 is unknown.
PCR based survey in S. Sofia of known Salmonella colonisation factors
As expected the SEF14, SEF18 and pef fimbrial gene sequences were not detected in S. Sofia. The agfA that encodes the major subunit of SEF17 fimbriae was found in all S. Sofia isolates.
Unexpectedly the lpfA gene for lpf (long polar fimbriae) was found in most S. Sofia isolates this is unexpected because the scientific literature suggests that this gene should not be present in a Salmonella enterica sub-species II organism of which S. Sofia is a member. The lpf fimbriae are predicted to be involved in the invasion of the Peyer's patches (lymphoid tissue) found in the mammalian and avian guts.
Knockout mutagenesis of colonisation factor genes and construction of isogenic mutants in S. Sofia
Cloning of the S. Sofia agfA and agfB genes which are members of the gene cluster for the synthesis of SEF 17 was achieved using PCR based upon primers to the equivalent sequence in S. Enteriditis.
DNA sequencing of these genes revealed few differences in the inferred S. Sofia proteins from those of S. Enteriditis and S. Typhimurium.
Since the agfA gene that encodes the major subunit of SEF17 appeared to be found universally in all S. Sofia tested it was decided to disrupt this gene by the insertion of an aphA3 kanamycin resistance cassette. This was achieved with relative ease. However, transfer of suicide vector with the mutated agfA gene into MH76 and recombination into the chromosome proved very difficult to achieve.
Eventually a number of mutant strains were isolated that were confirmed by PCR analysis.
Trial of S. Sofia colonisation factor mutants for colonisation of chickens
Due to the longer than expected time to produce MH76 mutants above, time did not permit the carrying out of this task.
Discussion
Conventional and Molecular typing
There has been no significant change in the distribution of serovars isolated from chickens with the exception that S. Virchow has become a consistent member of the most common serovars isolated from chickens and is now being isolated outside of its traditional niche in Queensland. The decline in numbers of S. Sofia from chickens appears to have been arrested. The Australian Salmonella Reference Centre has obtained an S. Virchow phage typing set, enabling phage typing to be carried out on this serovar. The predominant phage types found in humans and chickens in Australia are phage-types 8 and 34. Phage type 8 is relatively commonly reported elsewhere, however phage type 34 is relatively rare in the U.K.
The molecular typing methods that are used in our laboratory in conjunction with classical methods are based upon RFLP (Restriction Fragment Length Polymorphism) analysis. The examination of a wide variety of S. Virchow isolates using PFGE has demonstrated that this method can reveal differences within the common 8 and 34 phage types and also suggests a possible genetic relationship between these types. Using conventional phage typing it is impossible to obtain results in less than two full days; other methods such as PFGE (Pulsed Field Gel Electrophoresis) are even slower (4 days) and do not always yield a discriminatory result. This task was successful in that the classical and molecular typing methods that are in place are adequate in continuing to monitor Salmonella in Australian chicken flocks although there is a need for a more rapid method. The numbers of isolates sent by industry sources over the period of the grant would indicate satisfaction with the typing service.
Colonisation trials in Chickens with natural S. Sofia variants
This report has shown that some S. Sofia strains can colonise chickens for varying lengths of time up to 36 days. One strain, MH76 can successfully and consistently colonise one-day-old chickens when dosed orally. It had been observed that DNA sequences were variable around the invH gene in the SPI1 Pathogenicity Island in S. Sofia. PCR and Southern analysis had shown that a number of S.
Sofia isolates lacked the invH gene as well as adjacent genes in the inv cluster. The invH gene is thought to be a critical gene for the invasion of gut mucosa in calves, however, in chickens it is thought that fimbrial adhesins may be playing this role. A number of strains were tested were variable at invH with all having SEF17 and lpf fimbriae. Five strains were chosen for further study in the next task. In summary this task was completed successfully since a number S. Sofia strains capable of colonising chickens for extended periods were identified.
Challenge of S. Sofia strains selected for the ability to colonise chickens for long periods by virulent serovars.
S. Sofia strains that were capable of colonising chickens for long periods of time in the absence of virulent serovars such as S. Typhimurium, were tested in chickens in the presence of S. Typhimurium PT64 infected chickens. Strains MH7, MH89, MH107 and SSWT behaved in a similar fashion after contact with S. Typhimurium infected birds and were in most cases undetectable beyond day 12. In addition these birds became co-colonised with S. Typhimurium after contact on day 12 with S.
Typhimurium infected birds. The S. Typhimurium infected control birds remained colonised with that serovar until day 36. Strain MH76 persisted in the presence of S. Typhimurium, but it is impossible to conclude that the presence of MH76 is reducing S. Typhimurium numbers. It is clear that birds initially infected with S. Typhimurium can become co-infected with MH76 when placed in contact with birds initially infected with MH76. In conclusion, MH76 is a superior colonising strain of S. Sofia both alone and in the presence of S. Typhimurium. It is clear that S. Sofia does not exclude S. Typhimurium as both serovars can co-colonise regardless of initial colonisation status.
The MH76 strain however may be useful as a carrier for bacteriocin or other antibiotic expressing genes capable of reducing the colonisation of chickens with S. Typhimurium and other clinically significant serovars. In summary, this task was successfully completed. S. Sofia is not acting as a natural competitive agent in chickens against S. Typhimurium, however, some S. Sofia strains can co -colonise in the presence of S. Typhimurium and may be useful as carriers for beneficial genes.
PCR based survey in S. Sofia of known Salmonella colonisation factors
Examination of S. Sofia isolates by PCR, then confirmed by Southern analysis showed, that the sefA and sefD genes which encode SEF14 and SEF18 fimbriae respectively, were essentially absent from S. Sofia These fimbriae are found in S. Enteriditis as well as the host adapted serovars Gallinarum and Pullorum.. Likewise the plasmid encoded pef fimbriae were absent from all strains tested. The agfA gene which encodes the major subunit of the SEF17 fimbriae are predicted to occur on all Salmonella serovars and subspecies, which is not surprising since these fimbriae are thought to be involved in biofilm formation and binding to fibronectin. The presence of the lpfA gene (confirmed by sequence analysis) which encodes the major subunit of the long polar fimbriae (lpf) was an unexpected discovery in S. Sofia. The lpfA gene is predicted to exist only in sub-species I organisms, such as S. Typhimurium, S. Enteriditis etc. This task was successfully completed. In summary it confirmed that the pef, SEF14 and SEF18 pilin subunit genes are absent from S. Sofia as expected.
The universal presence of the agfA gene encoding the major subunit of SEF17 was confirmed. The lpfA gene was unexpectedly found and it is possible that the lpf fimbriae may be responsible for the continued reports of S. Sofia in chickens in Australia and elsewhere.
Knockout mutagenesis of colonisation factor genes and construction of isogenic mutants in S. Sofia
This task consisted of a number of sub-tasks; cloning of part of the agf gene cluster, insertional mutagenesis of the agfA gene with a kanamycin resistance cassette, DNA sequencing of the cloned agfA and agfB genes, and finally allelic exchange of the mutated agfA gene back into S. Sofia MH76.
DNA sequencing of the agfA/agfB cloned genes for the fimbrial subunit and an accessory protein respectively revealed that the inferred proteins from S. Sofia were very similar to those from S.
Enteriditis, S. Typhimurium and to a lesser extent E. coli. Disruption of the agfA gene was achieved by the insertion of an 800bp fragment of DNA encoding resistance to kanamycin. The final part of this task was to recombine the mutated agfA gene into the chromosome of MH76 and proved very difficult to achieve, but was ultimately successful. In summary, this task has been successful, as insertion mutants in agfA have been confirmed in MH76.
Trial of S. Sofia colonisation factor mutants for colonisation of chickens.
Not attempted due to lack of time.
Implications
Epidemiological and scientific evidence continues to support the notion that S. Sofia, which remains the most common Salmonella colonising chickens used for chicken meat production is of low virulence to humans. The discovery of a gene, lpfA, not predicted to be present in a sub-species II organism, like S. Sofia, is very interesting. These genes encode the long polar fimbriae (lpf) and have been implicated in the invasion of the lymphoid tissue in the gut (Peyer's patches) by sub- species I organisms. It could be hypothesised that the presence of the lpf genes may partially explain the continued colonisation of Australian chickens with S. Sofia. These findings do not explain why S. Sofia organisms can colonise chickens for extended periods of time in the presence of S. Typhimurium.
Recommendations
|
