Designing of polymerase chain reaction primers for the detection of Salmonella enteritidis in foods and faecal samples


  • S.-J. Wang,

    1. Department of Food Health, Chia Nan University of Pharmacy and Science, Tainan Hsien, Taiwan, ROC
    Search for more papers by this author
  • D.-B. Yeh

    1. Department of Food Health, Chia Nan University of Pharmacy and Science, Tainan Hsien, Taiwan, ROC
    Search for more papers by this author
    • Present address: Graduate Institute of Biotechnology, Chia Nan University of Pharmacy and Science, Tainan Hsien, 717 Taiwan, ROC.

S.-J. Wang Department of Food Health, Chia Nan University of Pharmacy and Science, Tainan Hsien, 717 Taiwan, ROC (


Aims: In this study, novel insertion element (IE) DNA targeted polymerase chain reaction (PCR) primers were designed and further used for the specific detection of Salmonella enteritidis in foods and faecal samples.

Methods and Results: Polymerase chain reaction primers, based upon their IE gene sequence (accession number Z83734), were developed for the detection of Salm. enteritidis. These primers were termed IE1L-IE1R and IE2L-IE3R. The cell lysate, rather than the extracted DNA, was used as template and preculturing of bacterial material was carried out prior to the PCR assay. The specificities of these developed primers were to be confirmed further. The PCR procedure developed was used to examine 170 endogenously contaminated samples, including poultry, seafood, meats, faecal specimens and some feed samples. Salmonella enteritidis was detected in 5·29% (nine of 170) of the samples.

Conclusions, Significance and Impact of the Study: Two sets of novel PCR primers, based upon their IE gene sequence, have been developed. These primers demonstrated the ability to be used for the specific detection of Salm. enteritidis. When PCR primers IE1L-IE1R were used for the detection of artificially Salm. enteritidis-contaminated food samples, as few as 1 cell ml−1 sample could be detected using this PCR process.


Salmonella is one of the most important pathogenic genera implicated in food-borne bacterial outbreaks and diseases (Gouws et al. 1998). Over the last few years, Salmonella enterica serotype enteritidis has appeared to be the most common serovar of Salmonella spp. in many countries (Usera et al. 1994; Landeras et al. 1996; Baylis et al. 2000). Most cases of salmonellosis occurred sporadically or as limited family or community outbreaks. However, some hospital- and nursing home-associated outbreaks have also been reported previously (Landeras et al. 1996). In addition, Salm. enteritidis has frequently been observed as a contaminant in foods such as eggs and poultry products (Cohen et al. 1994; Fadl et al. 1995; Landeras et al. 1996).

Traditional detection methods for Salmonella were based on cultures using selective media and followed by a series of biochemical and/or serological tests. As long as 5–7 d are generally required to confirm the presence of Salmonella species (Food and Drug Administration 1995). This time delay has led to the development of several alternative methods for the rapid detection of the presence of Salmonella sp., some of which have been commercialized (Gouws et al. 1998). Among these procedures, a series of DNA-based methods for the detection and identification of either all type serovars (Lin et al. 1996; Gouws et al. 1998; Tsen and Jian 1998) or specific species of Salmonella, such as typhimurium (Soumet et al. 1999a, 1999b), typhi (Song et al. 1993; Liu et al. 1995) and enteritidis (Cohen et al. 1994; Fadl et al. 1995; Frech and Schwarz 2000) have already been established recently. Various polymerase chain reaction (PCR)-based methods have also been published for the detection of enteritidis. For example, an arbitrarily primed PCR was developed to analyse the genomic DNA of enteritidis isolated from human outbreaks and an avian source (Fadl et al. 1995). However, enteritidis was detected in faeces from hens by using a PCR process. It was shown that these primers were specific to all members of the genus Salmonella, but not only exclusively to Salm. enteritidis (Cohen et al. 1994). A Multiplex PCR-based assay method with three sets of primers has been developed for the simultaneous detection of Salm. enteritidis, Salm. typhimurium and all Salmonella spp. (Soumet et al. 1999a, 1999b). However, unfortunately only relatively few serovars of Salmonella have been used to test the specificities of these primers.

In this study, two sets of novel PCR primers were developed based on their insertion element (IE) gene sequence; these primers demonstrate the ability to be used for the specific detection of Salm. enteritidis. The nucleotide sequences of these two primers appear to be characteristically different from other primers previously reported (Cohen et al. 1994; Soumet et al. 1999a, 1999b). In this paper, the approach to the design of new primers and the application of these primers in the PCR detection of Salm. enteritidis contamination in foods and faecal samples is described. For comparison, some of these samples were preinoculated with cultured bacterial cells of Salm. enteritidis.


Bacterial strains

The strains of Salm. enteritidis, other Salmonella sp. and non-Salmonella bacterial strains used in this study are listed in Table 1. These bacterial strains were obtained from multinational collaborators, including The American Type Culture Collection, MA, USA; The United States Department of Agriculture (USDA), Washington, DC, USA; Center for Disease Control, Atlanta, GA, USA; Culture Collection and Research Center, Hsin-Chu, Taiwan; National Laboratories of Food and Drugs, Taipei, Taiwan; Department of Health of the City of New York, NY, USA; the World Health Organisation; The Veterinary Service Laboratory of USDA, Ames, IA, USA; National Center for Disease Control, Taipei, Taiwan and the Pingtung University of Technology, Pingtung, Taiwan. One loopful of bacterial cells in Luria agar slants of the different species was inoculated into 5 ml Luria broth and cultured at 37°C for 8–12 h. Bacterial dilutions prepared with sterile water were used for all investigations in the PCR process.

Table 1.   Polymerase chain reaction results for the Salmonella and non-Salmonella isolates where primers IElL-IE1R and IE2L-IE3R were used Thumbnail image of

Designing of polymerase chain reaction primers

In this study, the PCR primers used for the detection of Salm. enteritidis included IE1L-IE1R and IE2L-IE3R (Table 2). The sequences, locations within genes and predicted sizes of the PCR products for these primers are also shown in Table 2. These primers were designed from an IE of Salm. enteritidis (accession number Z83734) by computerized sequence analysis.

Table 2.   Oligonucleotide primers used for the polymerase chain reaction (PCR) amplification of the genes of insertion element DNA for Salmonella enteritidisThumbnail image of

Polymerase chain reaction assay

For the PCR assay, the cell lysate was used as source material and the method of Victor et al. (1991) was modified for cell lysate preparation. In brief, 10 μl of the diluted heat-lysed cells were mixed with 30 μl PCR buffer containing 4 μl 10× PCR buffer (1× PCR buffer was 10 mmol l−1 Tris-HCl, pH 9·0, containing 50 mmol l−1 KCl, 1·5 mmol l−1 MgCl2 and 0·1% Triton X-100), each dNTP at a concentration of 0·25 mmol l−1, 50 pmol of each of the PCR primers and distilled H2O was added to a final volume of 40 μl. After heating at 95°C for 10 min, 2·5 units Taq DNA polymerase (Dynazyme; Finnzymes Oy, Espoo, Finland) in 10 μl 1× PCR buffer were added and the mixture subjected to PCR. Solvent evaporation was prevented by laying one drop of mineral oil (Sigma) on top of the reaction mixture. A microprocessor-controlled thermal cycler (Gene Amp PCR system 2400; Perkin Elmer, Norwalk, CT, USA) was used for automated temperature control during the PCR process. The DNA was denatured at 94°C for 2 min, followed by amplification through 35 cycles of denaturing, annealing and extension. Denaturation was performed at 94°C for 40 s, primer annealing being conducted at 55°C for 50 s and extension at 72°C for 50 s. Final extension was performed at 72°C for 2 min. The amplified product was detected by examination of a 10-μl aliquot of the PCR products by electrophoresis through a 1·8% ultra-pure agarose gel in 1× TBE buffer (5× TBE buffer contains 54 g Tris base, 27·5 g boric acid and 20 ml 0·5 mol l−1 EDTA, pH adjusted to 8·0 and made up to 1 l with distilled water). Identification of the resultant DNA bands was accomplished by comparison with 100-bp ladders of molecular weight markers after staining with ethidium bromide (Sigma).

Detection of Salmonella enteritidis in foods with a conventional bacteriological analytical manual and the developed polymerase chain reaction method

The food samples studied were purchased from local markets, feed stocks deriving from local poultry houses and faecal specimens acquired from patients with gastrointestinal disorders and healthy volunteers. A conventional method for the detection of Salmonella sp. as described in the Bacteriological Analytical Manual (BAM; Food and Drug Administration 1995) was used, further serotyping tests being used for the identification of Salm. enteritidis.


Specificity of the polymerase chain reaction primers

Upon analysis of the DNA sequences for genes encoded with an IE of Salm. enteritidis (accession number Z83734), Salm. typhimurium (X13616), Escherichia coli (J01733) or Shigella sp. (U97491), segments from a DNA sequence unique to Salm. enteritidis were selected and tested for their specificity according to the details of the testing procedure specified in Materials and Methods.

The oligonucleotide primer pairs used in this study were termed as IE1L-IE1R and IE2L-IE3R (Table 2). The sequences of these oligonucleotide primers were clearly different from those analogous reported previously by other workers, such as those designed from the virulent gene of Salm. enteritidis (Cohen et al. 1994; Soumet et al. 1999a, 1999b). Under the PCR conditions as described here, all of the 24 Salm. enteritidis strains generated PCR products with molecular weights of 316 and 559 bp using IE1L-IE1R and IE2L-IE3R as primer pairs, respectively (Table 1). Salmonella isolates other than enteritidis and bacterial strains other than Salmonella sp., including strains of the family Enterobacteriaceae such as E. coli, Shigella and Citrobacter, did not generate any false-positive results (Table 1). All of the Salm. enteritidis strains used in this study have been confirmed as being true using different PCR primers (Table 1) such as the primers reported by Soumet et al. (1999b), i.e. primers Sef167 and Sef478 which amplify the Sef14 fimbrial antigen, the virulence plasmid for Salm. enteritidis (Soumet et al. 1999b). It should be pointed out, however, that, for primers reported by Soumet et al. (1999b), only 14Salmonella serovar specimens have been used to test the relative specificities of these primers.

Detection sensitivity

Since an investigation of the PCR detection process after total DNA extraction by the phenol–chloroform method did not reveal sufficiently good results (data not shown), cell extracts prepared by heat lysis of the decimal serial dilution of the 8–12-h culture of Salm. enteritidis were used for the PCR assay. The detection sensitivity for Salm. enteritidis strains was 104 and 102 colony-forming units (cfu) ml−1 when IE2L-IE3R and IE1L-IE1R were used as primers, respectively (data not shown). The low sensitivity of the IE2L-IE3R pair as compared with the IE1L-IE1R pair may be due to the sequence differences between the primers which affect the primers' annealing efficiency with target DNA template. However, the detection sensitivity of the PCR process may also vary with the method used for target DNA preparation (Tsen et al. 1998) and the food samples selected for assay (Dickinson et al. 1995).

Detection of Salmonella enteritidis in various samples

Under the conditions described above, our results indicate that as few as 1–102 target cell(s) ml−1 in the sample could constitute sufficient cellular material to generate a positive PCR result for food and other samples following enrichment when IE1L-IE1R and IE2L-IE3R were used as primers, respectively (Fig. 1). From the data obtained, a preculturing step could improve the detection sensitivity of this technique.

Figure 1.

 Sensitivity of the detection of Salmonella enteritidis with IE1L-IE1R as primers in egg samples. Lanes: a, DNA ladder marker; b, polymerase chain reaction (PCR) result for blank without inoculation of the target cells; c–g, PCR results amplified from 10° to 105 cfu target cells ml−1 egg

Both PCR and BAM methods were used for the detection of the possible presence of Salm. enteritidis cells in unspiked foods and other samples (Table 3). The rate of contamination of Salm. enteritidis was detected as 5·29% by the developed PCR method using IE1L-IE1R as primers and as only 4·12% by the BAM method. The endogenously contaminating microflora in these samples constituted between 3 × 106 and 9 × 103 cfu g−1 of the various samples investigated (data not shown).

Table 3.   Detection of endogenous Salmonella enteritidis in foods and other samples Thumbnail image of

The results shown in Table 3 reveal that the BAM was less sensitive than the PCR method in detecting Salmonella. In the BAM method, the greater the number of isolated presumptive colonies from Salmonella–Shigella agar, the greater the frequency of accurate results (Shieh et al. 1996). Presumptive colonies must be isolated, otherwise false-negative results may occur.

For a large number of samples, PCR provides more rapid and efficient detection of food ingredients and end-products than the BAM method for bacterial contamination, thereby reducing possible warehouse and labour costs. In conclusion, this modified PCR assay method demonstrated good sensitivity and specificity as compared with the BAM culture method. Moreover, detection of Salm. enteritidis by this PCR method could be completed within 30 h as compared with the 5–7 d required for bacterial culture and a conventional serological method.


This study was financially supported by the Taiwan Sugar Cane Company and Wei-Li Pharmaceutical, Tainan, Taiwan, the Republic of China (Project No. S188003). The authors wish to express their gratitude to Dr Hau-Yang Tsen (National Chung-Hsing University) for his generous advice and providing most of the bacterial strains used in this study.


  1. Present address: Graduate Institute of Biotechnology, Chia Nan University of Pharmacy and Science, Tainan Hsien, 717 Taiwan, ROC.