**4.2.2 PFGE (pulsed field gel electrophoresis)**

356 Salmonella – A Dangerous Foodborne Pathogen

pattern other than those in the database, it was considered as reacting with the typing phages, but lytic pattern did not correspond to any recognized phage types, so called RDNC (= Reacting with the typing phage, but lytic pattern Did Not Correspond to any recognized phage types). But, we must note that phage typing analyses needs typing

In bref, phage typing can play an important role in surveillance and control of the common *Salmonella* serotypes. However, this requires strengthened efforts to make the system available to more laboratories internationally, possibly a simplification of the system to enhance its robustness even though this may slightly compromise its discriminatory power,

Phenotypic typing methods requiring enough time, personnel and reagent have led to the development of typing methods based on genotypic information. Currently used molecular typing methods are based on restriction endonuclease digestion, nucleic acid amplification,

Plasmid profile analysis was one of the earliest DNA-based subtyping schemes. It is particularly important, since most of the plasmids harbour virulence and antimicrobial resistance properties in *Salmonella*. Plasmid content of the host within the same serotype reveals the differentiation according to the profile (the number and molecular sizes of plasmids) obtained. The different plasmid profiles within a serotype points the lateral transfer by gaining or loosing the plasmid(s). The plasmids found in *Salmonella* differ in size

The detection method is based on the isolation of plasmids followed by agarose gel electrophoresis. Different protocols can be used (Helmuth et al., 1985). To view the plasmid pattern, agarose gel must be stained with ethidium bromide solution and then visualised

Plasmid analysis has several limitations. Plasmids can rapidly be acquired or lost. Also, single predominant plasmids have become endemic within various serotypes. In sporadic isolates of *S. enteritidis* from Maryland, 88% of isolates contained a single 36-Mda plasmid (Morris et al., 1992). Similarly, only 1 of 56 *S. typhimurium* isolates failed to encode a 90 kb plasmid, which is thought to be a serotype specific virulence plasmid. Despite the ubiquitous nature of the 90 kb plasmid, profiling of the entire complement of plasmids in each strain was able to discriminate *S. typhimurium* strains isolated from a single poultry

Plasmid analysis was also able to identify a multi-state outbreak of chloramphenicol resistant *S. newport* in humans that could be traced back to contaminated beef and to dairy farms (Riley et al.,1983). In a testament to the power of combining a strong traditional epidemiological analysis with serological and genotypic tests, a peak of *S. muenchen* was noted in Ohio, Michigan, Georgia and Alabama. Epidemiological studies failed to identify a common food source responsible for this outbreak, but a strong correlation with marijuana use was identified. Marijuana obtained from affected households was contaminated with *S. muenchen* and the isolates from the different states showed a similar plasmid fingerprint

suggesting interstate transfer of the contaminated drug (Taylor et al., 1982).

and finally improved external and internal quality assurance systems.

2 – 200 kb with different functionalities (Rychlik et al., 2006).

flock or closely related flocks (Millemann et al.,1995).

phage sets to be performed.

**4.2 Molecular methods** 

**4.2.1 Plasmid profiling** 

under UV light.

or nucleotide sequencing techniques.

PFGE has been considered as the "gold standard" among other molecular typing methods. By cutting the bacterial DNA with rare-cutting restriction endonucleases and running with special electrophoresis separation technique which use pulsed currents that change polarity at defined intervals, it separates the large fragments of DNA up to 12000 kb and yields strain specific patterns.

The choice of restriction endonuclease is somewhat empiric, but the most commonly used enzymes in *Salmonella* have been XbaI, SpeI and NotI. Comparisons of patterns from multiple enzymes can elucidate new subtypes and increase the discriminatory power of this technique (Liebisch & Schwarz, 1996).

PFGE of 60 *S. enteritidis* isolates revealed 28 different XbaI restriction profiles and 26 with SpeI, yet when the patterns generated from both enzymes were combined, 32 different pulsed-field types could be identified (Ridley et al., 1998). PFGE was used to determine whether molecular subtyping was able to detect unsuspected clusters or outbreaks of *S. typhimurium* (Bender et al., 2001). In fact, during a four-year period, 16% of isolates were linked to common source outbreaks. Of these, the authors felt that 62% of outbreak strains would have been missed without the use of PFGE molecular subtyping (Bender et al., 2001). PFGE has also been used to track outbreak strains occurring across national boundaries (Lyytikainen et al., 2000).

PFGE is characterized by a high degree of reproducibility both within and between laboratories (Swaminathan et al., 2001). The recent introduction of computerized gel-based data collection and analysis systems allows better standardization between laboratories thus creating the ability to rapidly compare restriction fragment patterns from isolates analyzed from remote locations (Swaminathan et al., 2001). Large databanks that house PFGE patterns from isolates around the world will greatly enhance *Salmonella* outbreak detection. PulseNet, a molecular subtyping network for foodborne bacterial disease surveillance, has been active in developing standardized PFGE protocols and establishing a national database. An outbreak of *S. agona* linked to contaminated cereal was identified in 1998. PFGE, in association with PulseNet, was used to identify cases in adjoining states that were not initially thought to be at risk (Swaminathan et al., 2001). In fact, combining typing methods such as PFGE and information from food chains, it was possible to identify related strains and common source of contamination. This type of approach may be useful in order to improve Salmonella spp. surveillance systems.

PFGE, however, is not always successful. Some serotypes, especially those with certain distinct phage types, can be so genetically homogeneous that multiple genotypic techniques fail to discriminate outbreak from non-outbreak strains. Ahmed et al. (Ahmed et al., 2000)

evaluated PFGE to differentiate *S. enteritidis* DT8 strains that developed during a Canadawide outbreak of gastroenteritis that was eventually traced to contaminated cheese. Successful discrimination was only achieved with a combination of intensive epidemiological, genotypic and phenotypic methods (Ahmed et al., 2000). Additionally, certain serotypes may be more susceptible to genetic rearrangements that can alter the PFGE pattern, even within an outbreak (Echeita & Usera, 1998).

Laboratory Typing Methods for Diagnostic of

**4.2.4 Insertion sequence (IS) typing** 

riboprinter itself.

Salmonella Strains, the "Old" Organism That Continued Challenges 359

electronically and the banding pattern from a particular organism can be compared to the entire databank stored in the computer. In contrast to PFGE, the time required to perform automated ribotyping is minimal; hybridization results can be obtained within 4 hours. A recent study tracking the rise of a multi-drug resistant, cephalosporin-resistant *S. newport*  proposes to use automated ribotyping as a way to rapidly identify the *newport* serotype and PFGE to further evaluate strain associations (Fontana et al., 2002). The major drawbacks of automated ribotyping are the high reagent costs per isolate and the cost of the automated

Laconha et al. and Ridley et al. investigated the genotypic differences between strains of *Salmonella* by plasmid analysis, ribotyping and pulsed-field gel electrophoresis (PFGE). The results obtained by those researchers indicated that PFGE may offer a better level of discrimination of *S. Enteritidis* types than other genotypic methods. Conversely, other epidemiological studies of *S. Enteritidis* have demonstrated that PFGE methodology has a

IS200 is a mobile element found in a variety of eubacterial genera, such as *Salmonella*, *Escherichia*, *Shigella*, *Vibrio, Enterococcus, Clostridium, Helicobacter,* and *Actinobacillus*. IS200 elements are very small (707-711 bp) and contain a single gene. Unlike typical mobile elements, IS200 transposes rarely. A consequence of IS200 self-restraint is that the number and distribution of IS200 elements remain fairly constant in natural populations of bacteria. This stability makes IS200 a suitable molecular marker for epidemiological and ecological studies, especially when the number of IS200 copies is high. IS200 typing, has been used to evaluate the molecular relationships between *Salmonella* isolates. In *Salmonella enterica*, IS200 fingerprinting is extensively used for strain discrimination. It is a 708 bp insertion sequence that is present in multiple copies within the *Salmonella* chromosome (Lam & Roth, 1983). Hybridization of digested chromosomal DNA with an IS200 probe has been useful in describing the clonal heritage of *Salmonella* from various serotypes, but has not been as discriminating as phage typing itself for *S. enteritidis*, *S. typhi* and others (Threlfall et al., 1994). For certain phage types of *S. typhimurium*, such as the multidrug resistant DT204c and 193 types common in the U.K., IS200 typing can result in strain discrimination and in some studies has been superior to PFGE and ribotyping (Jeoffreys et al., 2001). More frequently,

The standard RAPD technology (Williams et al., 1990) utilises short synthetic oligonucleotides (10 bases long) of random sequences as primers to amplify nanogram amounts of total genomic DNA under low annealing temperatures by PCR. Amplification products are generally separated on agarose gels and stained with ethidium bromide. Decamer primers are commercially available from various sources (e.g., Operon Technologies Inc., Alameda, California). PCR amplification with primers shorter than 10 nucleotides [DNA amplification fingerprinting (DAF)] has also been used producing more

Although these approaches are different with respect to the length of the random primers, amplification conditions and visualisation methods, they all differ from the standard PCR

lower discriminatory capacity than ribotyping (Olsen et al. 1994; Thong et al. 1998).

PFGE has performed better than IS200 typing (Amavisit et al., 2001).

complex DNA fingerprinting profiles (Caetano-Annoles et al., 1991).

**4.2.5 RAPD (randomly amplified polymorphic DNA)** 

Despite that PFGE is usually considered as the method of choice to determine the molecular relatedness among *Salmonella* strains; this method is relatively slow, often taking three days to complete, and requires the presence of expensive specialized equipment, high quality chemicals, and a considerable experience in the preparation of the DNA-containing agarose slices. Moreover, single genetic events, such as point mutations, integration, deletion or recombination events, can result in differences in the fragment patterns (Herschleb et al., 2007).
