Microbiology and Infectious Disease / SALMONELLA AND SALMONELLA SEROTYPE TYPHISPECIFIC PCR

Broad-Range (Pan) Salmonella and Salmonella Serotype TyphiSpecific Real-Time PCR Assays Potential Tools for the Clinical Microbiologist
John J. Farrell, MD,1,2 Laura J. Doyle, MT(ASCP),2 Rachel M. Addison, MPH, MT(ASCP),3 L. Barth Reller, MD,3 Geraldine S. Hall, PhD,2 and Gary W. Procop, MD2
Key Words: Pan-Salmonella; Salmonella Typhi; Real-time PCR
DOI: 10.1309/DP0HY5UT10HQW9YM

Abstract
We describe broad-range salmonellae (ie, Salmonella) and Salmonella serotype Typhispecific LightCycler (Roche Diagnostics, Indianapolis, IN) realtime polymerase chain reaction assays. We validated these with a battery of 280 bacteria, 108 of which were salmonellae representing 20 serotypes. In addition, 298 isolates from 170 clinical specimens that were suspected to possibly represent Salmonella were tested with the pan-Salmonella assay. Finally, the panSalmonella assay also was used to test DNA extracts from 101 archived, frozen stool specimens, 55 of which were culture-positive for salmonellae. Both assays were 100% sensitive and specific when cultured isolates of the battery were tested. The pan-Salmonella assay also characterized correctly all salmonellae on the primary isolation agar and was 96% sensitive (53/55) and 96% specific (49/51) when nucleic acid extracts from direct stool specimens were tested. These assays represent potential tools the clinical microbiologist could use to screen suspect isolates or stool specimens for Salmonella.

Salmonella species are important causes of enteritis throughout the world. In addition, Salmonella enterica serotypes Typhi and Paratyphi are important causes of enteric fever in underdeveloped countries that lack adequate sewage disposal and water treatment facilities.1,2 It is standard practice in many clinical microbiology laboratories to culture for Salmonella and Shigella species by using primary isolation media, such as MacConkey and Hektoen Enteric agars.3,4 Isolates suspected to possibly represent salmonellae (ie, lactose non-fermenting colonies with or without hydrogen sulfide production) are submitted for additional screening or identification, depending on individual laboratory practices.5-7 These practices, although effective, are timeconsuming and costly. The identification of Salmonella directly from the primary isolation plates or, ideally, directly from stool samples is attractive.8 This has been achieved using polymerase chain reaction (PCR) for the invA gene, a gene associated with the invasive nature of Salmonella.9 We hypothesized that the prgK gene, which is part of a genetic complex that also is thought to be important for enterocyte invasion, might serve as a useful target for a real-time PCR-based assay for the detection of Salmonella enterica from cultures and clinical stool specimens.10-12 In addition, we developed a real-time PCR assay directed against the vexC gene, which encodes for the Vi antigen, as a means of detecting the Salmonella serotype Typhi. The usefulness of the gene that encodes for the Vi antigen as a specific target for the identification of Salmonella Typhi has been demonstrated previously.13-15 The real-time PCR assays described were validated against a large battery of lysates from well-characterized bacteria, including 108 strains of Salmonella, representing 20 serotypes. In addition, we used
Am J Clin Pathol 2005;123:339-345
339
DOI: 10.1309/DP0HY5UT10HQW9YM

American Society for Clinical Pathology

339 339

Farrell et al / SALMONELLA AND SALMONELLA SEROTYPE TYPHISPECIFIC PCR

the pan-Salmonella real-time PCR assay to test culture isolates from the primary inoculation media of routinely obtained stool cultures, as well as DNA extracts from stool specimens that had been cultured.

Materials and Methods

Primer and Probe Design The primers and fluorescent resonance energy transfer (FRET) probes, for use with LightCycler system (Roche Diagnostics, Indianapolis, IN), for the broad-range or panSalmonella assay were designed based on the prgK gene of Salmonella serotype Typhimurium (GenBank accession number, AE008831).10,16 This primer set generated a 193-basepair nucleotide product that demonstrated high predicted specificity for Salmonella when submitted to a National Center for Biotechnology Information (NCBI) Blast, with perfect matches obtained only for Salmonella enterica. The GenBank accession numbers of these perfect matches were AE008831, L33855, AC117230, and AL627276. The PanSalm FRET hybridization probe set was designed to be 100% homologous for Salmonella Typhimurium. The 3' end of the donor FRET hybridization probe (PanSalmHP1) was labeled with fluorescein (fam). The 3' end of the acceptor FRET hybridization probe (PanSalmHP2) was phosphorylated (p) to prevent probe extension during PCR, and the 5' end was labeled with 705N-hydroxysuccinimide ester (LC Red 705). The sequences for the pan-Salmonella primers and probes were as follows: Pan-SalmF, 5'-CCTTTCTTATTGCGGGCA-3' (forward primer; position 4179-4196); PanSalmR, 5'-GCCGATGTGGATTATGAC-3' (reverse primer; position 4372-4355); PanSalmHP1, HP1 5'-GGATTGTTTTGATTATTTTGTTATCCGTGATG-[fam]-3' (donor FRET probe; position 4266-4235); and PanSalmHP2, 5'[LC Red 705]-AGCAGGCTTTGGCGTCTGGGTGG-p-3' (acceptor FRET probe; position 4232-4210). The primers and FRET hybridization probes used for the Salmonella Typhispecific real-time PCR assay were designed based on the vexC gene of Salmonella Typhi (GenBank accession number, AL627283), which encodes for the Vi antigen. This primer set generated a 215-base-pair nucleotide product that demonstrated high predicted specificity for Salmonella Typhi when submitted to an NCBI Blast, with perfect matches obtained only for Salmonella Typhi. The Salmonella Typhi GenBank accession numbers that demonstrated perfect matches were AL627283, AE016848, and D14156.1. The 3' end of the donor FRET hybridization probe (STyphiHP1) was labeled with fluorescein (fam). The 3' end of the acceptor FRET hybridization probe (STyphiHP2) was
340 340

phosphorylated (p) to prevent probe extension during PCR, and the 5' end was labeled with LC Red 640. The sequences for the Salmonella Typhispecific primers and probes were as follows: STyphiF, 5'-ACCCCGTAGCCCAATA-3' (forward primer; position 42395-42410); STyphiR, 5'-AGGAGAGACGCATTCG-3' (reverse primer; position 42610-42595); STyphiHP1, 5'-GCATATCGGTATTCTGGCGGC-[fam]-3' (donor FRET probe; position 42500-42480); and STyphiHP2, 5'-[LC Red 640]-CTGGTTCAGGCAAAACGACG-p-3' (acceptor FRET probe; position 42477-42458). PCR Reagents and Thermocycling Parameters LightCycler FastStart DNA Master Hybridization Probe kit reagents (Roche Diagnostics), primers, and probes were combined in a master mix. The PCR reaction was carried out in LightCycler capillary tubes (Roche Diagnostics) with 15 L of master mix and 5 L of sample. For the pan-Salmonella PCR assay, the following concentrations were used: 1.0 mol/L of both forward and reverse primers, 0.4 mol/L of both PanSalm FRET hybridization probes, and 4.0 mmol/L of magnesium chloride. The same concentrations were used for the Salmonella Typhispecific assay, except a 3.0-mmol/L concentration of magnesium chloride was used. The thermocycling protocol of the pan-Salmonella PCR consisted of 10 minutes at 95C for DNA polymerase activation, 45 cycles of PCR amplification (95C for 10 seconds, 50C for 10 seconds, and 72C for 20 seconds), postamplification melting (35C to 95C at 0.1C/second), and a cooling step (35C for 30 seconds). The protocol for the Salmonella Typhispecific assay was the same, except primer annealing occurred at 55C for 10 seconds and the postamplification melting ranged from 45C to 75C. Validation Battery The sensitivity and specificity of both assays were assessed using lysates of 280 well-characterized bacterial strains obtained from Duke University Medical Center (Durham, NC) and the Cleveland Clinic Foundation (Cleveland, OH) stock collections of bacteria. These bacteria included numerous American Type Culture Collection (Manassas, VA) strains. The other non-Salmonella stock isolates were identified by traditional biochemical methods by experienced microbiologists and were characterized unambiguously. Of the 280 isolates tested, 108 were Salmonella species, which included 23 Salmonella Typhi and 24 Salmonella Paratyphi. The 108 Salmonella isolates tested consisted of 20 serotypes, including those routinely identified (ie, commonly occurring) in these 2 clinical microbiology laboratories. The serotyping of these isolates was performed in the local public health laboratories according to routine. A variety of commonly encountered bacteria was included in the remaining 172 non-Salmonella bacteria tested and included
American Society for Clinical Pathology

Am J Clin Pathol 2005;123:339-345

DOI: 10.1309/DP0HY5UT10HQW9YM

Microbiology and Infectious Disease / ORIGINAL ARTICLE

other enteric pathogens such as Shigella species, Yersinia enterocolitica, and Escherichia coli O157:H7 Table 1. DNA lysates of the cultured bacteria were prepared by placing a 1.0-L loop of each cultured isolate into 500 L of lysis buffer, which has been described previously and contains 1 tris(hydroxymethyl)aminomethane-EDTA buffer supplemented with 1% Triton X-100 (Sigma, St Louis, MO) and 0.5% polysorbate 20.17 This solution was heated to 100C for 10 minutes, centrifuged at 10,000g for 30 seconds, and stored at 20C. The criterion for a positive test result was a cycle threshold (CT) (ie, crossing point) of 35 cycles or fewer. Limits of Detection The limit of detection was determined only for the panSalmonella assay because it was not our intention to use the Salmonella Typhispecific assay on direct specimens, but rather on cultured isolates wherein the target copy number is excessive. To determine the analytic sensitivity of this assay for lysates of Salmonella in pure culture, duplicate serial dilutions of a 0.5-McFarland suspension of Salmonella Typhimurium (American Type Culture Collection, 23854) were prepared in 0.85% saline. Lysates of each serial dilution were tested by the LightCycler prgK real-time PCR protocol as described. One hundred microliters from each of the serial dilutions was inoculated and spread onto sheep blood agar plates (Becton Dickinson Microbiology Systems, Cockeysville, MD) and then incubated in atmospheric oxygen at 35C for 24

hours to determine colony counts for each dilution. The lowest limit of detection was determined by averaging the values of the duplicate quantitative cultures that were correlated with the lowest limit of detection by PCR. A stool specimen, proven by culture to be negative for all bacterial enteric pathogens, was used to determine the analytic sensitivity of the pan-Salmonella assay for stool specimen extracts. Aliquots of this stool specimen were spiked with serial 10-fold dilutions of a 0.5-McFarland preparation of Salmonella Typhimurium, as described in the preceding text. DNA extracts were prepared, as described subsequently, for the clinical stool specimens, and real-time PCR was performed on these extracts and compared with the results of the quantitative cultures, which were performed as described in an earlier section. Testing Suspect Colonies From Routine Stool Cultures With the Pan-Salmonella Assay A prospective validation of the pan-Salmonella assay was conducted during a 5-month period on all isolates from stool cultures that were suspected to possibly represent Salmonella because of morphologic features of their colonies (ie, green/green or black colonies on Hektoen enteric agar). Cell lysates were prepared as described in the preceding text from 298 suspect isolates from the stool cultures of 170 patients. The lysates were batch tested once a week by the protocol described herein, with the same positive test criterion (ie, CT 35 cycles).

Table 1 Validation Battery of Cultured Bacteria*

PCR Results Bacteria Tested Pan-Salmonella Serotype Typhi Specific

Gram-positive bacteria tested (n = 40) Staphylococcus aureus (5), Staphylococcus epidermidis (3), Staphylococcus saprophyticus (2), All negative Micrococcus species (2), Stomatococcus species (2), Lactobacillus species (2), Enterococcus species (3), Viridans streptococci (3), Streptococcus pneumoniae (3), group A streptococcus (3), group B streptococcus (3), Aerococcus species (3), Listeria species (3), Bacillus species (3) NonSalmonella gram-negative bacteria tested (n = 132) Providencia species (2), Shigella sonnei (10), Shigella flexneri (17), Shigella boydii (6), Shigella All negative dysenteriae (2), Burkholderia species (2), Yersinia kristensenii (1), Yersinia enterocolitica (27), Citrobacter species (3), Escherichia coli non-O157:H7 (2), E coli O157:H7 (12), Proteus species (3), Klebsiella species (3), Enterobacter species (3), Pseudomonas species (3), Acinetobacter species (3), Haemophilus species (3), Neisseria meningitidis (3), Neisseria gonorrhoeae (3), Non-gonococcal Neisseria species (3), Moraxella species (3), Bacteroides species (3), Serratia species (3), Corynebacterium species (3), 1 each of Afipia felis, Vibrio cholerae, Eikenella corrodens, Pasteurella multocida, Campylobacter jejuni, Mesorhizobium species, Rhizobium species, Bartonella henselae, and Bartonella quintana Salmonella isolates tested (n = 108) Salmonella Typhimurium (15), Salmonella Enteritidis (6), Salmonella Choleraesuis (5), Salmonella All positive Paratyphi (24) Salmonella Infantis (8), Salmonella Heidelberg (5), Salmonella Arizonae (8), Salmonella Java (2), Salmonella Newport (2), and single isolates of Salmonella Agona, Salmonella Oslo, Salmonella Poona, Salmonella Senfrenberg, Salmonella Havana, Salmonella Javiana, Salmonella Anatum, Salmonella Saint Paul, Salmonella Berta, Salmonella Braenderup Salmonella Typhi (23) All positive
PCR, polymerase chain reaction. * Numbers in parentheses following names of bacteria indicate the number of strains tested.

All negative

All negative

All negative

All positive

American Society for Clinical Pathology 341

Am J Clin Pathol 2005;123:339-345

DOI: 10.1309/DP0HY5UT10HQW9YM

341 341

Farrell et al / SALMONELLA AND SALMONELLA SEROTYPE TYPHISPECIFIC PCR

Each batch included a DNA extract from Salmonella Typhimurium as a positive control sample and lysis buffer as a negative control sample. The assay results subsequently were compared with the culture results to determine the sensitivity and specificity of the assay in this setting. The Salmonella Typhispecific PCR was not performed on these colony extracts, primarily because of the low incidence of Salmonella Typhi in the United States and because the sensitivity and specificity of the Salmonella Typhispecific assay was defined previously in the validation phase of this study. DNA Extracts From Clinical Stool Specimens DNA extracts of 101 stool samples identified by culture or Campylobacter enzyme immunoassay to be positive for a bacterial enteric pathogen and 5 samples that were negative for enteric pathogens were tested. The pathogens detected by culture were Salmonella (55), Shigella (20), E coli O157:H7 (9), Campylobacter jejuni (4), Y enterocolitica (3), and Plesiomonas shigelloides (1). Nine stool samples were positive for Campylobacter by the ProSpecT (Alexon-Trend, Ramsey, MN) Campylobacter enzyme immunoassay. All stool samples were stored at 20C before extraction. Nucleic acid extracts were prepared from thawed stool specimens with the Qiagen Stool Extraction Kit (Qiagen, Valencia, CA). The pan-Salmonella realtime PCR assay was performed with the master mix and thermocycling protocol described in preceding text and with the same criteria for test positivity (ie, CT 35 cycles). The Salmonella Typhispecific PCR also was not performed on the direct stool extracts because of the low incidence of Salmonella Typhi in the United States. An evaluation of the Salmonella Typhi real-time

PCR on routinely occurring stool culture isolates and nucleic acid extracts from direct stool specimens should be performed in an area with a sufficient incidence of Salmonella Typhi infections to appropriately validate this assay.

Results
Validation Battery All lysates from the 108 Salmonella strains included in the validation battery were positive with the pan-Salmonella PCR, whereas only the 23 strains of Salmonella Typhi were positive with the Salmonella Typhispecific PCR (Table 1) Figure 1 and Figure 2. The 85 strains of Salmonella that were not serotype Typhi were negative with the Salmonella Typhispecific PCR. The 172 lysates from non-Salmonella bacteria were negative with both the pan-Salmonella and the Salmonella Typhispecific assays (Table 1). These assays both demonstrated 100% sensitivity and 100% specificity when the bacterial lysates from the validation battery were tested (Table 1). The approximate analytic sensitivity of the pan-Salmonella assay for cultured bacteria was 8 colonyforming units per milliliter. Routine Stool Culture Isolates All suspect colony lysates from the stool samples proven by culture to contain Salmonella in the prospective study were positive with the pan-Salmonella PCR assay. All remaining suspect colonies, determined to not represent salmonellae by

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.2

Fluorescence (F3/F1)

Fluorescence (F2/F1)

Positive control Negative control Non-Salmonella Salmonella

2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.2

Positive control Negative control E coli O157:H7 Non-Typhi Salmonella

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 Cycle Number

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 Cycle Number

Figure 1 Pan-Salmonella real-time polymerase chain reaction. The positive control (Salmonella Typhimurium) and all Salmonella isolates (16 strains, 12 serotypes) in this experiment demonstrates unequivocal amplification compared with the negative control sample (lysis buffer) and the non-Salmonella isolates (Yersinia species [3 strains] and Escherichia coli O157:H7 [11 strains]) tested. (If no number of strains is given, 1 strain was assayed.)

Figure 2 The Salmonella serotype Typhispecific real-time polymerase chain reaction demonstrated amplification of only the Salmonella Typhi isolate (ie, the positive control). The negative control sample (lysis buffer), non-Salmonella bacteria (Escherichia coli O157:H7 [11 strains]), and non-Typhi Salmonella isolates (19 strains, 15 serotypes) were all negative.

342 342

Am J Clin Pathol 2005;123:339-345

DOI: 10.1309/DP0HY5UT10HQW9YM

American Society for Clinical Pathology

Microbiology and Infectious Disease / ORIGINAL ARTICLE

routine bacteriologic methods, were negative with the panSalmonella PCR. During this study, there was 1 case of apparent misinoculation of lysis buffer or a mislabeling of the test tube that resulted in an apparent false-positive result; reevaluation of the 2 colony types involved gave correct identification with the pan-Salmonella assay. Therefore, as with the validation battery, we believe the sensitivity and specificity of the pan-Salmonella PCR on colony lysates from colonies on the primary recovery media were both 100%. PCR Testing of Frozen, Archived Stool Specimens Of the 55 stool specimens positive by culture for salmonellae, 53 were positive with the pan-Salmonella PCR (ie, there were 2 false-negative PCR results) Table 2. Of the extracts from the 51 stool specimens that were culture negative for salmonellae, there were 2 that tested positive in the pan-Salmonella assay (ie, 2 false-positive results). Both were from stool samples from which E coli O157:H7 had been cultured. The sensitivity and specificity for the pan-Salmonella assay when DNA extracts from stool specimens were tested were 96% (53/55) and 96% (49/51), respectively (Table 2). Low-level amplification (CT = 44) was detected in one of the stool extracts, but this had to be categorized as a false-negative result because of our predefined interpretation criteria. The testing of DNA extracts of normal stool samples spiked with serial dilutions of Salmonella Typhimurium to determine the limit of detection of this assay demonstrated that PCR was positive at a dilution of 105 but not at greater dilutions. Duplicate quantitative cultures of 100 L of this dilution revealed 91 organisms. Therefore, the limit of detection of this assay for Salmonella in DNA extracts of stool specimens is estimated at approximately 910 colony-forming units per milliliter.

Table 2 PCR of Stool Specimen Extracts

Pan-Salmonella PCR Result Stool Culture/EIA Results Total Salmonella-positive stool samples Salmonella serotype Typhi O antigen serogroups Group B Group C Group D nonSalmonella Typhi Group E or G Not typeable Other enteric pathogens and negative stool cultures Campylobacter jejuni Campylobacter by EIA Shigella sonnei Shigella flexneri Yersinia enterocolitica Escherichia coli O157:H7 Plesiomonas shigelloides No pathogen isolated N 55 1 20 15 14 2 3 51 4 9 19 1 3 9 1 5 Positive Negative 53 1 19 14 14 2 3 2 0 0 0 0 0 2 0 0 2 0 1 1 0 0 0 49 4 9 19 1 3 7 1 5

EIA, enzyme immunoassay; PCR, polymerase chain reaction.

Discussion
The taxonomy of Salmonella has been reviewed extensively by Janda and Abbott.18 Two species of Salmonella currently are recognized, S enterica and Salmonella bongori. Eight subgroups or subspecies have been defined using a variety of molecular methods.18-20 S bongori contains only a single subspecies or subgroup (ie, subgroup V, bongori). S enterica contains the remaining 7 subgroups, 2 of which contain the clinically relevant salmonellae. Most of these are contained within subgroup I, enterica, and are subcategorized further as serotypes (eg, Typhi, Typhimurium, Enteritidis).21,22 Subgroup IIIa, arizonae, contains the clinically important salmonellae that are associated with reptiles and cause human infection. We designed and tested and now report a sensitive and specific LightCycler real-time PCR assay for the detection of Salmonella and another for the detection of the Typhi serotype.
American Society for Clinical Pathology

The pan-Salmonella assay was directed against a single-copy gene, which is part of a complex that is thought to be important for the entry of salmonellae into enterocytes.23,24 A similar genetic target, the invA gene, previously has been shown to be useful for the detection of salmonellae by PCR.25,26 We think this assay potentially could be useful as a molecular means of screening colonies from stool cultures that have characteristics that suggest the possibility of salmonellae. We proved this hypothesis, first by testing lysates from a large battery of wellcharacterized bacteria, many of which are isolated commonly in a clinical microbiology laboratory. The validation battery also included 108 strains of Salmonella, representing 20 serotypes, including Typhi, Paratyphi, and commonly occurring serotypes. The pan-Salmonella and the Salmonella Typhispecific assay correctly characterized all isolate lysates as a Salmonella species or the Typhi serotype, respectively. The additional studies performed examined only the panSalmonella assay, which was the assay that was of greatest interest to us for implementation into our laboratory for routine testing. Furthermore, we thought additional studies of the Salmonella Typhispecific PCR assay on stool culture isolates or stool nucleic acid extracts from patients in Cleveland would not sufficiently validate this assay further, given the low incidence of Salmonella Typhi infections in this area. These studies, which are important before assay implementation, should be performed in an area wherein Salmonella Typhi is endemic to validate the assay. We further examined the pan-Salmonella assay in a prospective study wherein we examined a large number of lysates (n = 298) from colonies suspected to possibly represent
Am J Clin Pathol 2005;123:339-345
343
DOI: 10.1309/DP0HY5UT10HQW9YM

343 343

Farrell et al / SALMONELLA AND SALMONELLA SEROTYPE TYPHISPECIFIC PCR

Salmonella because of their features on the primary inoculation media. These 298 isolates were from the routine stool cultures of 170 patients. All colonies were categorized appropriately as Salmonella or non-Salmonella species based on the results of the pan-Salmonella assay. These results largely were expected, given the results of the testing of the validation battery. This aspect of the study, however, was done on lysates prepared by routine medical technologists and done directly from media used for primary inoculation. There was a misinoculation or a mislabeling of a lysate, which reiterates the need for good quality control with any laboratory test. Although not performed in this study, the possibility exists of testing multiple different colonies or sweeps from individual plates, rather than individual colonies. The feasibility of this approach remains to be proven, but it likely is possible given the low limit of detection of this assay for lysates derived from cultured isolates. Although the Salmonella Typhispecific assay was not tested on the colonies that grew on the primary inoculation media because of the low incidence of enteric fever in the United States, it is highly likely that this assay could be used to exclude the Typhi serotype for isolates that tested positive within the pan-Salmonella assay; this, too, however, remains speculative. This prospective validation at the stool bench in our clinical laboratory demonstrated that real-time PCR was as accurate for the differentiation of Salmonella from non-Salmonella isolates as conventional methods, but considerably faster. Not surprisingly, the pan-Salmonella assay was considerably less sensitive (approximately 2 logs) when DNA extracts from stool specimens were tested compared with colony lysates. However, even with extracts from this complex specimen type, the assay was highly sensitive with 96% (53/55) of Salmonella-containing stool samples detected. Two of the 51 stool samples that did not contain salmonellae by culture were positive with the pan-Salmonella PCR. Both of the extracts that generated false-positive results were from cultures that grew E coli O157:H7. There was no cross-reactivity with this assay, and the E coli O157:H7 isolates tested (Figure 1), nor is there any expected based on GenBank searches. The possibility remains that salmonellae were present in these stool samples but not detected by routine culture (ie, false-negative cultures), but this remains speculative. The greater than 95% sensitivity and specificity of this assay for detecting Salmonella in extracts of stool specimens suggests that, perhaps owing to the large number of organisms typically present in stool samples from patients with Salmonella enterocolitis, this assay has adequate sensitivity for a clinical application. The complete Salmonella genome contains only 1 copy of the prgK gene, which might limit the analytic sensitivity of this assay and, thus, the applicability of this assay for the detection of Salmonella directly in the blood of patients with enteric fever, in whom the number of bacteria per milliliter would be considerably lower than in stool samples.
344 344

Initial studies (data not shown) included the use of an alternative acceptor FRET probe, the sequence of which was: 5'-[LC Red 705]-AGCAGGCTTTGGCGTCTGG-p-3' (acceptor FRET probe; position 4232-4210). This probe, however, did not detect Salmonella arizonae. This is not surprising, in retrospect, because S arizonae is in a distinctly different subgroup (IIIa) from the other clinically important salmonellae (subgroup I). Unfortunately, the sequence of the prgK gene for S arizonae was not available in GenBank for electronic analysis when this assay was designed. When we discovered that S arizonae was missed by the initial version of the pan-Salmonella assay, we first demonstrated that the amplicon was generated by the primers, through gel electrophoresis, and then sequenced the amplicons of S arizonae, Salmonella Typhimurium, and Salmonella Typhi for comparison. This comparison led to the redesign of the acceptor FRET hybridization probe used in this study. The extension of the 3' of the acceptor FRET probe by 4 nucleotides (GTGG) was necessary to detect S arizonae. The use of PCR and, more recently, real-time PCR has become commonplace in molecular microbiology and molecular pathology laboratories. This technology often is used to detect fastidious microbes or microorganisms that cannot be cultivated. Real-time applications, however, have been used for the detection of commonly cultured bacteria and their mechanisms of resistance. Examples of this include the detection of Staphylococcus aureus and the mecA gene and the detection of vancomycin-resistant enterococci.27,28 In some cases, real-time PCR has been found to be cost-competitive compared with culture.29 The detection of Salmonella and the differentiation of these organisms from the many mimics that occur in stool cultures are now possible with real-time PCR of lysates from cultured colonies or direct nucleic extracts of the stool specimens. How such technologies will be incorporated into the stool bench of the future, however, remains to be seen.
From the 1Department of Infectious Diseases, University of Illinois/OSF Infectious Diseases, Peoria, IL; 2Section of Clinical Microbiology, Department of Clinical Pathology, Cleveland Clinic Foundation, Cleveland, OH; 3Clinical Microbiology Laboratory, Duke University Medical Center, Durham, NC. Address reprint requests to Dr Procop: Clinical Microbiology, Dept of Microbiology L40, Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195.

References
1. Bern C, Martines J, de Zoysa I, et al. The magnitude of the global problem of diarrhoeal disease: a ten-year update. Bull World Health Organ. 1992;70:705-714. 2. Parry CM, Hien TT, Dougan G, et al. Medical progress: typhoid fever. N Engl J Med. 2002;347:1770-1782. 3. Bopp CA, Brenner FW, Fields PI, et al. Escherichia, Shigella, and Salmonella. In: Manual of Clinical Microbiology. 8th ed. Washington, DC: ASM Press; 2003:chap 42.

Am J Clin Pathol 2005;123:339-345

DOI: 10.1309/DP0HY5UT10HQW9YM

American Society for Clinical Pathology

Microbiology and Infectious Disease / ORIGINAL ARTICLE

4. Procop G. Gastrointestinal infections. Infect Dis Clin North Am. 2001;15:1073-1108. 5. Bonev S, Zakhariev Z, Gentchev P. Comparative study of media for determination of lysine decarboxylase activity. Appl Microbiol. 1974;27:464-468. 6. Kolmos HJ, Schmidt J. Failure to detect hydrogen-sulphide production in lactose/sucrose-fermenting Enterobacteriaceae, using triple sugar iron agar. Acta Pathol Microbiol Immunol Scand. 1987;95:85-87. 7. Johnson JG, Kunz L, Barron W, et al. Biochemical differentiation of the Enterobacteriaceae with the aid of lysine-iron-agar. Appl Microbiol. 1966;14:212-217. 8. Kurowski PB, Traub-Dargatz JL, Morley PS, et al. Detection of Salmonella spp in fecal specimens by use of real-time polymerase chain reaction assay. Am J Vet Res. 2002;63:12651268. 9. Rahn K, De Grandis SA, Clarke RC, et al. Amplification of an invA gene sequence of Salmonella typhimurium by polymerase chain reaction as a specific method of detection of Salmonella. Mol Cell Probes. 1992;6:271-279. 10. Pegues DA, Hantman MJ, Behlau I, et al. PhoP/PhoQ transcriptional repression of Salmonella typhimurium invasion genes: evidence for a role in protein secretion. Mol Microbiol. 1995;17:169-181. 11. Jones B, Falkow S. Identification and characterization of a Salmonella typhimurium oxygen-related gene required for bacterial internalization. Infect Immun. 1994;62:3745-3752. 12. Klein JR, Fahlen TF, Jones BD. Transcriptional organization and function of invasion genes within Salmonella enterica serovar Typhimurium pathogenicity island 1, including the prgH, prgI, prgJ, prgK, orgA, orgB, and orgC genes. Infect Immun. 2000;68:3368-3376. 13. Hashimoto Y, Itho Y, Fujinaga Y, et al. Development of nested PCR based on the ViaB sequence to detect Salmonella typhi. J Clin Microbiol. 1995;33:775-777. 14. Hirose K, Itoh K, Nakajima H, et al. Selective amplification of tyv (rfbE), prt (rfbS), viaB, and fliC genes by multiplex PCR for identification of Salmonella enterica serovars Typhi and Paratyphi A. J Clin Microbiol. 2002;40:633-636. 15. Sharma K, Arya S. Detection of Salmonella typhi by nested PCR based on the ViaB sequence [letter]. J Clin Microbiol. 1995;33:3361. 16. McClelland M, Sanderson KE, Spieth J, et al. Complete genome sequence of Salmonella enterica serovar Typhimurium LT2. Nature. 2001;413:852-856. 17. Reischl U, Linde HJ, Metz M, et al. Rapid identification of methicillin-resistant Staphylococcus aureus and simultaneous species confirmation using real-time fluorescence PCR. J Clin Microbiol. 2000;38:2429-2433.

18. Janda J, Abbott S. The Enterobacteria. Philadelphia, PA: Lippincott-Raven; 1998. 19. Brown EW, Kotewicz ML, Cebula TA. Detection of recombination among Salmonella enterica strains using the incongruence length difference test. Mol Phylogenet Evol. 2002;24:102-120. 20. Chan K, Baker S, Kim CC, et al. Genomic comparison of Salmonella enterica serovars and Salmonella bongori by use of an S enterica serovar Typhimurium DNA microarray. J Bacteriol. 2003;185:553-563. 21. Boyd EF, Wang F, Whittam TS, et al. Molecular genetic relationships of the salmonellae. Appl Environ Microbiol. 1996;62:804-808. 22. Koneman EW, Allen SD, Janda WM, et al. Color Atlas and Textbook of Diagnostic Microbiology. 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 1997. 23. Kubori T, Sukhan A, Aizawa SI, et al. Molecular characterization and assembly of the needle complex of the Salmonella typhimurium type III protein secretion system. Proc Natl Acad Sci U S A. 2000;97:10225-10230. 24. Pucciarelli M, Garcia-del Portillo F. Protein-peptidoglycan interactions modulate the assembly of the needle complex in the Salmonella invasionassociated type III secretion system. Mol Microbiol. 2003;48:573-585. 25. Isacsson J, Cao H, Ohlsson L, et al. Rapid and specific detection of PCR products using light-up probes. Mol Cell Probes. 2000;14:321-328. 26. Scholz HC, Arnold T, Marg H, et al. Improvement of an invA-based PCR for the specific detection of Salmonella typhimurium in organs of pigs. Berl Munch Tierarztl Wochenschr. 2001;114:401-403. 27. Shrestha N, Tuohy M, Hall G, et al. Rapid identification of Staphylococcus aureus and the mecA gene from BacT/ALERT blood culture bottles by using the LightCycler system. J Clin Microbiol. 2002;40:2659-2661. 28. Sloan L, Uhl J, Vetter E, et al. Comparison of the Roche LightCycler vanA/vanB detection assay and culture for detection of vancomycin-resistant enterococci from perianal swabs. J Clin Microbiol. 2004;42:2636-2643. 29. Shrestha N, Shermock K, Gordon S, et al. Predictive value and cost-effectiveness analysis of a rapid polymerase chain reaction for preoperative detection of nasal carriage of Staphylococcus aureus. Infect Control Hosp Epidemiol. 2003;24:327-333.

American Society for Clinical Pathology 345

Am J Clin Pathol 2005;123:339-345

DOI: 10.1309/DP0HY5UT10HQW9YM

345 345