Iris E. Valentin-Bon
Robert E. Brackett
Peter S. Holt
U.S. Food and Drug Administration
Center for Food Safety and Applied Nutrition, HFS-300
College Park, MD
U.S.A.
Introduction
Salmonella serotype Enteritidis (SE) emerged as a major cause of human salmonellosis world wide throughout the 1980-1990s and remains an important serovar. Of the 309 SE outbreaks with a confirmed source, 241 (78%) were associated with raw or undercooked shell eggs between 1990-2001 (2).
Grade-A shell eggs have been attributed to the human SE-related illness because SE can infect ovarian tissues of laying hens and be deposited into the developing egg, resulting in contamination of egg contents. Studies indicate that the frequency of infected eggs is extremely low and intermittent with most contaminated eggs having fewer than 20 cells per egg (9). Therefore, current implemented culture methods recommend incubation of pools of egg contents at room temperature for a minimum of 3 to 5 days prior to plating on culture media to improve the rate of detection (12).
Conventional means for identification and enumeration of SE in shell eggs are laborious and time consuming. It generally requires steps of enrichment, biological test, and serotyping which requires to 5-7 days to complete. We compared the effectiveness of the direct plating method recommended by the USDA/APHIS with a preenrichment procedure for the detection of SE in shell eggs. Homogenization using a stomacher, an electric blender, and hand massaging has been adopted in various studies for SE detection in eggs (3, 7, 8). Mixing egg contents with a sterile spoon or other sterile instrument is recommended in bacteriological analytical manual (BAM) methods (13). However, to date, the effect of homogenization methods on the growth of SE in liquid whole egg has never been reported. We compared the multiplication of SE in liquid whole egg pools homogenized with four different methods: mechanical stomaching, electric blending, hand massaging, and hand stirring, and determine the optimum homogenization method for isolating SE from raw eggs.
Supplementing pools of egg contents with iron in the form of ferrous sulphate and concentrated enrichment broth has been suggested to improve detection of SE from raw eggs without using enrichment broth (4, 5, 6). Hence, a technology that can speed up the detection and enumeration of SE in eggs would be valuable to egg industries and regulatory agencies. The efficacy of iron supplementation on the recovery of Salmonella Enteritidis (SE) in white large shell eggs using Ferroxiamine E (FE) was determined.
Rapid methodologies such as ELISA and PCR techniques for SE detection in eggs have been developed to overcome the drawbacks of conventional methods (8, 10, 11, 14). In contrast to conventional methods, these tests can detect SE in two days. However, they are not free of drawbacks. The tests involve time-consuming enrichment incubations, exhibit varying degrees of cross-reactions and both systems have been known to produce false positive reactions. The sensitivity of rapid detection methods for SE in eggs was substantially decreased due to the interference of egg contents (1). Cudjoe et al. (4) found that the more viscous, undiluted mixtures of eggs showed the highest bead loss compared with diluted samples when immunomagnetic beads were applied to recover SE from raw eggs. We have been collaborating with Neogen Corporation, Lansing, Michigan, to develop an antibody-based lateral flow assay (LFA) system to rapidly detect Group D Salmonellae from poultry samples. The current study was undertaken to evaluate a simple and rapid lateral flow device for its ability to detect SE (a group D) and not other Salmonella serovars from eggs. We also examined the factors that either enhance or decrease the detection capability of the test panel. Direct detection of SE from egg without enrichment broth using a novel antigen extraction method is also described.
Novel 5' nuclease assays (TaqMan assay) have been used for rapid detection of food borne pathogens such as Listeria monocytogenes, Salmonella, and E. coli O157:H7 (3, 10, 11). The 5' nuclease assay uses the 5'→ 3' nuclease activity of Taq DNA polymerase, which digests an internal fluorogenic probe resulting in an increase in reporter fluorescence signal that can be detected on a fluorescence spectrometer in a real time manner. Therefore, the increase in the reporter dye fluorescence is a direct consequence of target amplification. The point at which the fluorescence emission crosses a predetermined value is reported as the threshold cycle (Ct) number and is proportional to the initial number of the target template. The detection as well as quantification of specific target organism can be achieved by analyzing the Ct number. Among other molecular-based detection technologies, a real time PCR could fulfil the requirement for its practical application by food analysis laboratories. We developed a real-time PCR method for the detection of SE in shell eggs in which the serotype specific target sequence was the SEF14 gene. The specificity and sensitivity of the developed method were tested. The performance of the real-time PCR was evaluated by comparing with the conventional culture method using artificially inoculated egg pools with very low numbers of SE cells.
Materials and methods
1. Preenrichment Versus Direct Selective Agar Plating for the detection of Salmonella Enteritidis in shell eggs:
Preenrichment method
From each of the labelled pools, 25 ml test portions were added to 225 ml portions of trypticase soy broth supplemented with ferrous sulphate (mTSB; 35 mg ferrous sulphate/liter broth) and mixed by swirling. T
he test portions were allowed to equilibrate for 1 h, and the pH was adjusted to 6.8 with 1 N HCl or 1 N NaOH.
Preenrichments were incubated for 24h at 35°C. Aliquots of 1 ml were subcultured from the incubated broths to 10 ml portions of tetrathionate broth (TT). Aliquots (0.1 ml) from the same test portions were transferred to 10 ml portions of Rappaport-Vassiliadis (RV) medium. TT broth cultures were incubated for 24 hours at 35°C, and RV broth cultures were incubated for 24 hours at 42°C in a thermostatically controlled water bath (LabLine, AquaBath, Chicago, Ill.). Incubated selective enrichments were streaked to plates of bismuth sulfite (BS), brilliant green with and without 0.02 mg/liter of novobiocin (BGN and BG, respectively), xylose lysine desoxycholate (XLD), xylose lysine tergitol 4 (XLT4) agars, and incubated for 24 hours at 35°C. BS agar plates were incubated for an additional 24 hours (48 hours total). Presumptive colonies were picked and inoculated into triple sugar iron agar (TSI) and lysine iron agar (LIA) slants and incubated for 24 hours at 35°C. Isolates from presumptive-positive TSI slants were tested for agglutination with Group D1 somatic antiserum. All dehydrated media and sera were obtained form Difco Laboratories.
Direct agar plating method (APHIS method)
From each of the three bulk samples of eggs inoculated to contain three different levels of inocula (100 to 103 CFU/ml), the same 20 pools (600 to 650 ml of jumbo eggs, 400 to 500 ml of medium eggs) used in the preenrichment method were directly streaked on BGN, BG, XLD, and XLT4 agar plates for isolation and incubated up to 42 hours at 37°C. Presumptive SE colonies were picked and transferred to TSI and LIA slants and incubated for 24 hours at 35°C. Isolates from TSI slants were confirmed with Group D1 somatic antiserum.
2. Comparison of sample preparation methods for recovering Salmonella Enteritidis from eggs:
Sample homogenization
Stomacher: The 4 pools/inoculum were stomached for 30 seconds using a Stomacher lab blender 400® (Seward Laboratory, London, England) at normal speed setting (200 rpm).
Hand massage: The 4 pools/inoculum in sterile stomacher bags (Tekmar Company) were homogenized using hand massage for 30 seconds to ensure complete mixing of albumen and egg yolk material.
Hand stirring: The 4 pools/inoculum were aseptically poured into a 1000 ml sterile beaker and then egg yolks and egg whites were thoroughly mixed with a sterile spoon for 30 seconds. The mixed egg contents were transferred back into the stomacher bag for incubation.
Blending with electric blender: The 4 pools/inoculum were aseptically poured into a sterile blender jar and blended for 30 seconds at normal speed (Model 31BL91, Waring Commercial Laboratory, New Hartford, Conn). The blended egg contents were transferred back into the stomacher bag for incubation.
3. Selective isolation of Salmonella Enteritidis from mixed cultures in egg using ferrioxamine E (FE) and ferrous sulphate (FS) supplementation:
Preparation of inoculated eggs
The air-dried eggs were aseptically broken and then the yolk and albumen were separated into sterile plastic 200 ml beaker. Whole egg contents and egg whites were stomached for 30 seconds using a stomacher lab blender 400 (Seward Laboratory, London, England) at normal speed. Sixteen of 9 ml of the homogenized samples per each group were allocated into 15 ml conical test tubes and divided into two groups, supplemented and unsupplemented. Each group was inoculated with 0.1 ml E. coli K12 101 to 109 CFU/ml. The supplemented group was supplemented with FS or FE with optimum concentration. All samples were inoculated with 0.1 ml of ca. 100 CFU/ml SERR and incubated 24 hours until sampled.
4. Simple and rapid methods for detecting Salmonella Enteritidis in raw eggs using Immunochromatograph Assay:
Egg samples
Egg samples were prepared as explained in experiment 1. From each of the labelled pools, 25 ml test portions were added to 225 ml portions of TSB, mTSB, RV, and TT for single enrichment samples. Double enrichment samples were prepared transferring 1 ml and 0.1 ml of mTSB into 9 ml of TT and 9.9 ml of RV broth, respectively. A loopful of samples from all different enrichment groups including direct egg samples was streaked on BGN agar plates.
Lateral Flow Assay (LFA)
A portion (100µl) of samples from the same group of enrichment broth including direct egg samples was placed into the round sample port of the test device, initiating a lateral flow through a reagent zone containing specific anti-Salmonella Enteritidis antibodies conjugated to colloidal gold particles. The antibody used in the device was developed in our lab and described in a previous study (Holt et. al., 1995). Test results were interpreted as positive or negative, scored on a scale from 0 (-) to 3 (+++) in proportion to the colour intensity determined visually using SE pure culture, after 20 min from the addition of the sample culture (Figure 1).
5. Rapid, specific detection of Salmonella Enteritidis in pooled eggs by real time PCR:
Oligonucleotides:
The PCR oligonucleotide primers and probe were designed according to the published DNA sequence of the SEF14 gene (GenBank accession #: L03833) in order to amplify a chromosomal DNA sequence of 98 bp (Table 1).
The oligonucleotides and all reagents for PCR reaction used in this study were synthesized and purchased from Applied Biosystems (Foster City, Calif.). The SE probe was labelled with 6-carboxyfluorescein (FAM, the reporter dye) and 6-carboxytetramethylrhodamine (TAMRA, the quencher dye).
Extraction and preparation of DNA templates for PCR
Four sets of each sample (1 ml) were centrifuged at 16000 x g for 3 min. The cell pellets were resuspended in PrepManTM Ultra and placed in a 100C water bath for 10 min. The samples were cooled for 2 min at room temperature and centrifuged. The supernatant fluids (5 ul) were transferred to 45 ul of PCR reaction mix consisting of TaqMan Universal PCR Master Mix (25 ul), forward primer (5 ul), reverse primer (5 ul), TaqMan probe (5 ul), and water (5 ul) in a 96-well microwell plate. The microwell plates were sealed with optical adhesive covers and were placed in an ABI-Prism 7000 sequence detector (Applied Biosystems). The reaction was run at 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 15 seconds and 60°C for 60 s. The total assay time was approximately 2 hours. The fluorescence measurements were taken in real-time and at the end were analyzed by the SDS software installed on the sequence detector.
Results and discussion
1. Preenrichment Versus Direct Selective Agar Plating for the detection of Salmonella Enteritidis in shell eggs:
The total number of positive test portions obtained with the preenrichment method was significantly higher (P<0.05) than obtained with the USDA/APHIS direct plating method (Table 1).
2. Comparison of sample preparation methods for recovering Salmonella Enteritidis from eggs:
The effect of different homogenization methods on the SE growth in egg pools inoculated with 10 SE CFU/pool is shown in Table 2. After 24 hours incubation at 37°C, the SE population recovered from samples prepared by manual means (hand-massaging or hand stirring) was approximately 106 CFU/ml. However, no SE in samples prepared by mechanical means (stomaching or electric blender) was detected after 24 hours incubation. Following a further 24 hours incubation, the population of SE in samples prepared by hand massaging or spoon hand stirring was approximately 3.9 x 108 CFU/ml, whereas the population of cells in SE positive samples (75%) prepared by stomaching was 3.1 x 106 CFU/ml after 48 hours incubation. No SE growth was observed in samples homogenized using electric blending. The current study clearly shows that the homogenization method is a critical factor for improving rapid detection of SE in eggs.
3. Selective isolation of Salmonella Enteritidis from mixed cultures in egg using ferrioxamine E (FE) and ferrous sulphate (FS) supplementation:
The growth of SE in the whole egg contents supplemented with either FS or FE was greater than E. coli growth up to 10 cells SE: 105 cells E. coli ratio (Table 3). Dramatic growth difference was observed between the supplemented group and the unsupplemented group. The most significant selective growth was shown in egg white samples supplemented with FE when the mixed culture ratio is lower than 100 (Table 4).
4. Simple and rapid methods for detecting Salmonella Enteritidis in raw eggs using Immunochromatograph Assay:
The minimum concentration of SE to generate a positive band on the test panel was approximately 107 cell/ml in pure culture (Figure 1) and no cross-reactivity was detected with two other Salmonella serovars, ST and SK in pure culture (Data not shown).
The flow lateral test kits were evaluated and compared with a direct plating method using bulk egg samples artificially contaminated with various concentration of SE in 4 trials. Four trials of an experiment were performed using SE phage type 13a to compare a conventional culture method and the flow lateral test kits assay. In all trials, all uninoculated negative egg pools were SE negative when tested by conventional culture method or immuno assay (Data not shown).
The immuno chromatograph assay detected equivalent or less number of egg pools containing SE than the traditional culture method with the exception of mTSB or TSB enrichment broths inoculated with 20 CFU in trial 1 (Table 5). However, no significant difference between two methods was observed (P>0.05) with the exception of mTSB enrichment broth inoculated with 17 CFU in trial 2. The number of pools yielding SE were increased in samples where enrichment steps were involved, regardless of what method used (Table 5). With the test kits, more positive samples were detected than direct plating when old eggs were used for double enrichments combining mTSB and TBG broths in trial 4.
5. Rapid, specific detection of Salmonella Enteritidis in pooled eggs by real time PCR:
Comparison of conventional culture method and real time PCR
Two trials of an experiment were performed using SE phage type 13a in trial 1 and SE phage type 4 in trial 2 to compare a conventional culture method and the real time PCR assay. In both trials, all uninoculated negative egg pools were SE negative when tested by conventional culture method or real time PCR assay (Data not shown). The developed real time PCR assay detected equivalent or more number of egg pools containing SE than the traditional culture method with the exception of direct egg pools inoculated with 51 CFU in trial 1 and 53 CFU/200 eggs in trial 2 (Table 6). However, no significant difference between two methods was observed (P>0.05). The number of pools yielding SE were increased in samples where enrichment steps were involved, regardless of what method used (Table 6). With the real time PCR assay, only a single enrichment was required to detect equivalent or more pools containing SE when compared with results from the conventional culture method (Table 6).
Conclusion
In summary, conventional cultural methods for detecting Salmonella Enteritidis in eggs and environmental samples are generally laborious and time consuming. However, the FDA's Bacteriological Analytical Manual (BAM) specifies preenrichment, selective enrichment, and selective plating for the recovery of Salmonella spp. in foods to allow for the recovery of injured cells and detection of low numbers of cells.
The preenrichment method for detection of SE in eggs was superior to the direct plating method. The low contamination rate and the antimicrobial effect of egg albumen have contributed to the difficulty in detecting SE in raw eggs. Therefore, homogenization of egg contents is an essential step for SE isolation from the egg contents. In all three experiments, whether the initial inoculum level was low or high, recovery of SE from eggs that were homogenized by mechanical means, notably by electric blender, was not as good as from eggs homogenized by manual means. Supplementing pools of egg contents with iron in the form of ferrous sulphate such as FS or FE and concentrated enrichment broth could improve detection of SE from raw eggs without using enrichment broth. The antibody-based test kits have high specificity, take only 30 minutes to finish, very reliable, and easy to use. However, it still requires double enrichments steps to achieve high sensitivity equivalent to conventional culture methods.
The current study showed that a simple and rapid test panel kit could be successfully used to detect SE in eggs in conjunction with various enrichment methods and lead to saving time and labour.
We developed a real-time quantitative PCR assay that enabled us:
1) to discriminate SE among 43 other Salmonella or non-Salmonella strains,
2) to detect approximately 102 CFU/ml in PBS and 103 CFU/ml in raw egg,
3) to show potential for quantifying SE in eggs,
4) to detect very low numbers of SE cells in egg pools in conjunction with a single enrichment.
The real time PCR assay proved to be a rapid, highly sensitive test for detection and quantification of low levels of SE in egg samples. Although the practical application of real time PCR is feasible in regulatory laboratory, the fact that the instrumentation is expensive and costly to operate may hinder routine testing at the farm level.
Acknowledgments
The authors thank Lindsay Chen, Grace Thammasuvimol, Brian Eblen, Antonio De Jesus, and Sharon Edelson Mammel for their tireless assistance and thoughtful comments.
References
1. Brigmon, R.L., Zam, S.G., Wilson, H.R., 1995.Detection of Salmonella Enteritidisin eggs and chicken with enzyme-linked immunosorbent assay. Poultry Science74:1232-1236.
2. CDC, 2003. Outbreaks of Salmonellaserotype Enteritidis infectionassociated with eating raw or undercooked shell eggs-United States, 1999-2001. MMWR51:1149-1152.
3. Chen, H., R. C. Anantheswaran, and S. J.Knabel. 2001. Optimizationof iron supplementation for enhanced detection of Salmonella Enteritidis in eggs. J. FoodProt. 64:1279-1285.
4. Cudjoe et al., 1994 Cudjoe,K.S., Krona, R., Grøn, B., Olsen, E., 1994. Use of ferrous sulphate and immunomagneticseparation to recover Salmonella Enteritidis from raw eggs.International Journal of Food Microbiology 23, 149-158.
5. Gast and Beard, 1992 Gast, R.K., Beard, C.W., 1992. Detection andenumeration of Salmonella Enteritidis in fresh and stored eggs laid byexperimentally infected hens. Journal of Food Protection 55:152-156.
6. Gast and Holt, 1998 Gast R.K, Holt, P.S., 1998. Supplementing poolsof egg contents with broth culture media to improve rapid detection of SalmonellaEnteritidis. Journal of Food Protection 61, 107-109.
7. Gast, R. K. 1993. Recovery of Salmonella Enteritidis frominoculated pools of egg contents. J. Food. Prot. 56:21-24.
8. Holt, P.S., R. K. Gast, and C.R. Greene. 1995. Rapid Detection of SalmonellaEnteritidis in pooled liquid egg samples using a magnetic bead-ELISAsystem. J. Food Prot. 58:967-972.
9. Humphrey, T.J., 1994. Contamination of egg shell and contents with SalmonellaEnteritidis: a review. Int. J. Food Microbiol. 21:31-40.
10. McElroy, A. P., N. D. Cohen, and B. M. Hargis. 1996. Evaluation ofthe polymerase chain reaction for the detection of Salmonella Enteritidis inExperimentally inoculated eggs and eggs from experimentally challenged hens. J.Food Prot. 59: 1273-1278.
11. Rijpens, N., H. Lieve, F. Vereecken, G. Jannes, J. De Smedt, and L.De Zutter. 1999. Rapid detection of stressed Salmonella spp. in dairyand egg products using immunomagnetic separation and PCR. Int. J. FoodMicrobiol. 46: 37-44., 2020.
12. U.S. Department of agriculture. 1993. Proposed rules: chickendisease caused by Salmonella Enteritidis.Fed. Regist. 58:41048-41061.
13. U.S. Foods and Drug Administration. 2002. Bacteriological analyticalmanual. AOAC International , Gaithersburg, Md. Chap. 5:Salmonella.Available at: http://www.cfsan.fda.gov/~ebam/bam-toc.html.Accessed 19 September 2002.
14. Woodward, M. J., and S. E. S. Kirwan. 1996. Detection of SalmonellaEnteritidis in eggs by the polymerase chain reaction. Vet. Rec.138:411-413.
From Proceedings of the"Midwest Poultry Federation Convention", St. Paul, Minnesota, U.S.A.
