Abstract
Most Salmonella serotypes alternatively express either phase-1 or phase-2 flagellar antigens, encoded by the fliC and fljB genes, respectively. Flagellar phase reversal for the identification of both flagellar antigens is not necessary at the genetic level. Variable internal regions of the fliC genes encoding the H:i, H:r, H:l,v, H:e,h, H:z10, H:b, and H:d antigens have been sequenced; and the specific sites for each antigen in selected Salmonella serotypes have been determined. These results, together with flagellar G-complex variable internal sequences obtained by the Foodborne and Diarrheal Diseases Branch at the Centers for Disease Control and Prevention in Atlanta, Ga., have been used to design a multiplex PCR to identify the G-complex antigens as well as the H:i, H:r, H:l,v, H:e,h, Hz10, H:b, and H:d first-phase antigens. These antigens are part of the most common Salmonella serotypes possessing first-phase flagellar antigens. Salmonella enterica serotype Enteritidis is identified by adding a specific primer pair published previously (P. G. Agron, R. L. Walker, H. Kinde, S. J. Sawyer, D. C. Hayes, J. Wollard, and G. L. Andersen, Appl. Environ. Microbiol. 67:4984-4991, 2001). This multiplex PCR includes 13 primers. A total of 161 Salmonella strains associated with 72 different serotypes were tested. Each strain generated one first-phase-specific antigen fragment ranging from 100 to 500 bp; Salmonella serotype Enteritidis, however, generated two amplicons of 500 bp that corresponded to the G complex and a 333-bp serotype-specific amplicon, respectively. Twenty-three strains representing 19 serotypes with flagellar genes different from those targeted in this work did not generate any fragments. The method is quick, specific, and reproducible and is independent of the phase expressed by the bacteria when they are tested.
Salmonella enterica is believed to be responsible for the majority of cases of food-borne disease worldwide (14). Serotyping is widely used as an epidemiological typing method to divide Salmonella strains into groups on the basis of surface antigenic variability. Phage typing, pulsed-field gel electrophoresis, and antibiogram analysis can then be used for further subdivision of the strains within serotypes. In most countries, Salmonella isolates are typed by the Kauffman-White (K-W) scheme. A serotype of Salmonella is determined on the basis of antigenic variability at lipopolysaccharide moieties (O antigen), the flagellar protein (H antigen), and the capsular polysaccharide (Vi antigen). To date, 2,501 Salmonella serotypes have been identified (15). Salmonella isolates are diphasic with respect to the flagellar antigens, with several monophasic exceptions (4). Salmonella is unique among the members of the family Enterobacteriaceae, as it commonly has two distinct flagellar antigens, phase 1 and phase 2, that are coordinately regulated so that only one flagellar antigen is expressed at any time. In Spain, only 10 serotypes, all of subspecies I, represent 89.8% of the clinical isolates studied in the Spanish National Reference Laboratory for Salmonella and Shigella (LNRSS). Serotypes Typhimurium and Enteritidis alone represent 73.9% of the clinical isolates (16).
Although serotyping offers a reliable method for differentiating Salmonella strains, identification by the slide agglutination method with a complete set of sera is a time-consuming process that requires the use of 167 specific antisera and well-trained technicians. At present, alternative methods for the identification of serotypes, such as DNA-based serotyping, or “molecular serotyping,” are under development in many laboratories around the world, including LNRSS (5, 6, 12). These molecular methods are highly sensitive, very specific, and reproducible. These methods also allow for better laboratory-to-laboratory quality control, as they are fairly standard in most laboratories. Molecular serotyping schemes that are based on the K-W Salmonella serotyping scheme allow the continuation of surveillance data analysis based on this scheme. Thus, the information that has been collected over many decades could not only be maintained but also continue to be collected.
The fliC and fljB genes encode the phase-1 and phase-2 flagellins, respectively. These genes are found at two different locations on the chromosome. The fljBA operon contains hin, which encodes the Hin recombinase; the fljB gene, which encodes the phase-2 flagellin; and the fljA gene, which encodes a repressor for the fliC gene. The Hin recombinase gene catalyzes the reversible inversion of a 993-bp segment of the chromosome containing a promoter for the fljBA operon. In one orientation, the promoter directs the transcription of the fljB and fljA genes, inducing repression of the fliC gene. In the other orientation of hin, fljB and fljA are not expressed, the phase-2 flagellin is turned off, and fliC is derepressed, allowing phase-1 flagellin to be expressed (9, 19). Some of these alleles are defined by a single factor (antigen i, d, or r); others are defined by several subfactors (e.g., antigens l,v, g,m, and e,n,x) (15). Comparison of the amino acid sequences of Salmonella flagellins has led to the definition of eight variable regions. The amino- and carboxy-terminal sequences (regions I and II and region VIII, respectively) are conserved and are thought to be important for polymerization and transportation. The central region, which comprises regions IV, V, and VI, is highly variable in both sequence and length between flagellar antigen genes and is generally believed to determine the epitope of the H antigen (13, 18).
In the past year, LNRSS has developed a PCR method that can identify second-phase flagellar antigens (complex H1, H:l,w, H:e,n,x, and H:e,n,z15) belonging to the most common serotypes isolated in Spain (5, 6). The aim of the project described here was to develop a multiplex PCR method that can detect the most common phase-1 flagellar alleles by an approach similar to that developed for the phase-2 flagellar alleles described previously (5, 6).
MATERIALS AND METHODS
Strains.
The fliC gene was sequenced from 47 strains belonging to 27 different serotypes of S. enterica subsp. enterica and expressing H:i, H:r, H:l,v, H:e,h, H:z10, H:b, and H:d first-phase flagellar antigens (Table 1). For the multiplex PCR assay, 137 strains expressing the antigens listed above were used as positive controls and 23 strains expressing H:a, H:c, H:k, H:y, H:z, H:z29, H:z38, H:z4,z23, H:z4,z24, and H:z4,z32 were used as negative controls (Table 2). All strains were from LNRSS, except for serotypes Rhone (Centers for Diseases Control and Prevention, Atlanta, Ga.); Tallase (Central Public Health Laboratory, London, United Kingdom); and Arechavaleta, Budapest, Chester, Kamoru, Kiel, Moscow, Naestved, Neumuenster, Pensacola, Rostock, and Wayne (World Health Organization International Laboratory for Serotyping, Pasteur Institute, Paris, France). Sequences from G-complex alleles were obtained from the Centers for Diseases Control and Prevention. These included the sequences of the alleles H:f,g (serotype Derby), H:f,g,s (serotype Agona), H:g,m and H:g,m,p (serotype Enteritidis), H:g,m,p,s (serotype Montevideo), H:g,m,s,t (serotypes Namib and Mosselbay), H:g,m,t (serotype Pensacola), H:g,p (serotype Dublin), H:g,p,s (serotype Naestved), H:g,p,u (serotype Rostock), H:g,q (serotype Moscow), H:g,s,t (serotypes Missouri and Senftenberg), H:g,t (serotypes Agodi and Merseyside), and H:m,t (serotype Oranienburg).
TABLE 1.
Antigenic formulas for the Salmonella strains used to sequence the first-phase flagellar antigens
| Serogroup | Serotype | Antigenic formula | Na | Consensus sequence | GenBank accession no. |
|---|---|---|---|---|---|
| C-1 | Livingstone | 6,7,14:d:l,w | 2 | ICSd-2 | AY434707 |
| E-1 | Madjorio | 3,10:d:e,n,z15 | 1 | ICSd-1 | AY434706 |
| G | Tilburg | 1,3,19:d:l,w | 2 | ICSd-3 | AY434708 |
| B | Typhimurium | 1,4,[5],12:i:1,2 | 2 | ICSi-1 | AY429608 |
| C-1 | Garoli | 6,7:i:1,6 | 1 | ICSi-1 | AY429609 |
| C-2 | Kentucky | 8,20:i:z6 | 1 | ICSi-2 | AY429612 |
| C-2 | Poeseldorf | 8,20,54:i:z6 | 1 | ICSi-2 | AY429613 |
| F | Aberdeen | 11:i:1,2 | 1 | ICSi-1 | AY429610 |
| X | Bergen | 47:i:1,5 | 1 | ICSi-1 | AY429611 |
| C-1 | Mbandaka | 6,7,14:z10:e,n,z15 | 2 | ICSz10-1 | AY434692 |
| C-2 | Hadar | 6,8:z10:e,n,x | 4 | ICSz10-2 | AY434693 |
| E-1 | Lexington | 3,10[15][15,34]:z10: 1,5 | 2 | ICSz10-3 | AY434694 |
| B | Wien | 1,4,12,27:b:lw | 3 | ICSb-2 | AY434697 |
| C-1 | Ohio | 6,7,14:b:lw | 3 | ICSb-1 | AY434695 |
| N | Urbana | 30.b:e,n,x | 1 | ICSb-1 | AY434696 |
| B | Saintpaul | 1,4,[5],12:e,h:1,2 | 1 | ICSe,h-2 | AY434699 |
| C-1 | Braenderup | 6,7,14:e,h:e,n,z15 | 2 | ICSe,h-1 | AY434698 |
| E-1 | Anatum | 3,10[15][15,34]:e,h:1,6 | 3 | ICSe,h-3 | AY434700 |
| B | Heidelberg | 1,4,[5],12:r.1,2 | 1 | ICSr-1 | AY434701 |
| C-1 | Infantis | 6,7,14:r:1,5 | 2 | ICSr-1 | AY434702 |
| C-1 | Virchow | 6,7,14:r:1,2 | 3 | ICSr-2 | AY434705 |
| C-2 | Bovismorbificans | 6,8,20:ri:1,5 | 1 | ICSr-1 | AY434703 |
| C-2 | Goldcoast | 6,8:r:lw | 1 | ICSr-1 | AY434704 |
| B | Brandenburg | 4,[5],12:l,v:e,n,z15 | 2 | ICSl,v-1 | AY434709 |
| B | Bredeney | 1,4,12,27:l,v:1,7 | 2 | ICSl,v-1 | AY434710 |
| D-1 | Panama | 1,9,12:l,v:1,5 | 1 | ICSl,v-2 | AY434711 |
| E-1 | Give | 3,10[15],[15,34]:[d]l,v:1,7 | 1 | ICSl,v-3 | AY434712 |
N, number of strains sequenced.
TABLE 2.
Antigenic formulas for the Salmonella strains used for multiplex PCR identification of the first-phase flagellar antigens
| Serogroup | Serotype | Antigenic formula | Na | Amplicon size (bp) | Serogroup | Serotype | Antigenic formula | Na | Amplicon size (bp) | |
|---|---|---|---|---|---|---|---|---|---|---|
| A | Kiel | 1,2,12:g,p:− | 1 | 500 | ||||||
| B | Derby | 4,[5],12:f,g:1,2 | 2 | 500 | ||||||
| B | Agona | 4,12:f,g,s:− | 3 | 500 | ||||||
| B | California | 4,12:g,m,t:− | 1 | 500 | ||||||
| C-1 | Rissen | 6,7:f,g:− | 4 | 500 | ||||||
| C-1 | Montevideo | 6,7:g,m,s:− | 3 | 500 | ||||||
| C-1 | Oranienburg | 6,7,14:m,t:[z57] | 1 | 500 | ||||||
| D-1 | Berta | 1,9,12:[f],g,[t]:− | 1 | 500 | ||||||
| D-1 | Budapest | 1,4,12,27:g,t:− | 1 | 500 | ||||||
| D-1 | Moscow | 1,9,12:g,q:− | 1 | 500 | ||||||
| D-1 | Naestved | 1,9,12:g,p,s:− | 1 | 500 | ||||||
| D-1 | Rostock | 1,9,12:g,p,u:− | 1 | 500 | ||||||
| D-1 | Pensacola | 9,12:m,t:− | 1 | 500 | ||||||
| E-4 | Senftenberg | 1,3,19:g,s,t:− | 2 | 500 | ||||||
| N | Wayne | 30:gz51:− | 1 | 500 | ||||||
| D-1 | Enteritidis | 9,12:g,m:− | 7 | 500 and 333 | ||||||
| C-2 | Hadar | 6,8:z10:e,n,x | 11 | 400 | ||||||
| C-1 | Mbandaka | 6,7,14:z10:e,n,z15 | 3 | 400 | ||||||
| E-1 | Lexington | 3,10,[15][15,25]:z10:1,5 | 2 | 400 | ||||||
| H | Nessa | 1,6,14,25:z10:1,2 | 1 | 400 | ||||||
| B | Brandenburg | 4,[5],12:l,v:e,n,z15 | 5 | 330 | ||||||
| B | Bredeney | 1,4,12,27:l,v:1,7 | 4 | 330 | ||||||
| D-1 | Panama | 1,9,12:l,v:1,5 | 7 | 330 | ||||||
| D-1 | Kapemba | 9,12:l,v:1,7 | 1 | 330 | ||||||
| E-1 | Give | 3,10[15][15,34]:[d],l,v:1,7 | 3 | 330 | ||||||
| D-1 | London | 3,10[15]:l,v:1,6 | 2 | 330 | ||||||
| C-1 | Infantis | 6,7,14:r:1,5 | 3 | 275 | ||||||
| C-1 | Virchow | 6,7,14:r:1,2 | 5 | 275 | ||||||
| E-1 | Weltevreden | 3,10[15]:r:z6 | 1 | 275 | ||||||
| F | Rubislaw | 11:r:e,n,x | 1 | 275 | ||||||
| B | Typhimurium | 4,12:i:1,2 | 8 | 250 | ||||||
| F | Veneciana | 11:i:e,n,x | 4 | 250 | ||||||
| C-2 | Kentucky | 8,20:i:z6 | 2 | 250 | ||||||
| B | Chester | 1,4,5,12:e,h:e,n,x | 1 | 200 | ||||||
| C-1 | Braenderup | 6,7,14:e,h:e,n,z15 | 3 | 200 | ||||||
| E-1 | Anatum | 3,10[15][15,34]:e,h:1,6 | 6 | 200 | ||||||
| C-2 | Newport | 6,8,20:e,h:1,2 | 4 | 200 | ||||||
| C-2 | Kottbus | 6,8:e,h:1,5 | 1 | 200 | ||||||
| B | Abony | 1,4,[5],12,27:b:e,n,x | 1 | 150 | ||||||
| B | Wien | 1,4,12,27:b:l,w | 2 | 150 | ||||||
| B | Paratyphi B | 1,4,[5],12:b:1,2 | 1 | 150 | ||||||
| B | Paratyphi B var. Java | 1,4,[5],12:b:1,2 | 1 | 150 | ||||||
| B | Wagenia | 1,4,12,27:b:e,n,z15 | 4 | 150 | ||||||
| C-1 | Ohio | 6,7,14:b:l,w | 4 | 150 | ||||||
| E-4 | Tambacounda | 1,3,19:b:e,n,x | 1 | 150 | ||||||
| N | Urbana | 30:b:e,n,x | 1 | 150 | ||||||
| C-1 | Livingstone | 6,7,14:d:l,w | 2 | 100 | ||||||
| C-2 | Manhattan | 6,8:d:1,5 | 2 | 100 | ||||||
| C-2 | Muenchen | 6,8:d:1,2 | 3 | 100 | ||||||
| E-1 | Madjorio | 3,10:d:e,n,z15 | 1 | 100 | ||||||
| E-4 | Tilburg | 1,3,19:d:l,w | 2 | 100 | ||||||
| D-1 | Ndolo | 9,12:d:1,5 | 1 | 100 | ||||||
| G | Grumpensis | 12,23:d:1,7 | 2 | 100 | ||||||
| B | Miami | 1,9,12:a:1,5 | 1 | —b | ||||||
| B | Sendai | 1,9,12:a:1,5 | 1 | — | ||||||
| B | Arechavaleta | 4,[5],12:a:1,7 | 1 | — | ||||||
| B | Tinda | 1,4,12,27:a:e,n,z15 | 1 | — | ||||||
| B | Neumuenster | 1,4,12,27:k:1,6 | 1 | — | ||||||
| B | Kamoru | 1,4,12,27:y:z6 | 1 | — | ||||||
| B | Coeln | 1,4,[5],12:y:1,2 | 1 | — | ||||||
| C-1 | Mikawasima | 6,7,14:y:e,n,z15 | 3 | — | ||||||
| C-1 | Lille | 6,7,14:z38:− | 1 | — | ||||||
| C-1 | Thompson | 6,7,14:k:1,5 | 2 | — | ||||||
| C-1 | Choleraesuis | 6,7:c:1,5 | 1 | — | ||||||
| C-1 | Tennessee | 6,7,14:z29:[1,2,7] | 1 | — | ||||||
| C-2 | Albany | 8,20:z4,z24:− | 1 | — | ||||||
| C-2 | Tallahassee | 6,8:z4,z32:− | 1 | — | ||||||
| G | Worthington | 1,13,27:z:l,w | 2 | — | ||||||
| K | Cerro | 6,14,18:c:z4,z23:[1,5] | 1 | — | ||||||
| L | Rhone | 21:c:e,n,x | 1 | — | ||||||
| M | Nima | 28:y:1,5 | 1 | — | ||||||
| O:51 | Treforest | 1,51:z:1,6 | 1 | — |
N, number of strains.
—, no PCR products were obtained.
DNA sequencing.
The fliC genes from the 47 strains shown in Table 1 were amplified by PCR with primer 1 and primer 2, as described previously (8). PCR amplification was performed with Ready-To-Go PCR beads (Amersham Biosciences, Piscataway, N.J.). Five microliters of a strain suspension that had been boiled for 10 min and briefly centrifuged was used as the template. PCR amplification was performed as follows: denaturation at 95°C for 5 min; 30 cycles of 95°C for 1 min, 58°C for 1 min, and 72°C for 2 min; and a final extension cycle at 72°C for 7 min. The amplification products were purified and sequenced by using primers sense-60 (17) and Reverse-fliC (ACTTCGGTTTTGCCGTCTGCGCC). Sequencing was performed on an ABI PRISM 377 DNA sequencer (Applied Biosystems, Applera Hispania, S.A., Barcelona, Spain) and the Taq Dye Deoxy Terminator cycle sequencing kit (Applied Biosystems/Perkin-Elmer).
Other sequences from GenBank were also used. These included sequences with GenBank accession numbers Z15064 (serotype Berta fliCf,g,t), Z15065 (serotype Budapest fliCg,t), Z15066 (serotype Derby fliCf,g), Z15067(serotype Dublin fliCg,p), D78639 (serotype Naestved fliCg,p,s), Z15068 (serotype Enteritidis fliCg,m), Z15069 (serotype Montevideo fliCg,m,s), Z15070 (serotype Oranienburg fliCm,t), Z15071(serotype Rostock fliCg,p,u), Z15072 (serotype Senftenberg fliCg,s,t), and Z15086 (serotype Moscow fliCg,q) (10).
Sequence analysis.
Sequence analysis was performed with Lasergene (version 5.0) software (DNA-Star, Madison, Wis.). The variable internal region (VIR) sequences of each allele were aligned and compared to determine the internal consensus sequence (ICS). To avoid cross-reactions among the antigens belonging to the same complex, i.e., H:E (e,h, e,n,x, and e,n,z15) and H:L (l,v and l,w), the sequences of previously identified alleles fljBenz15, fljBenx, fljBlw, and fliClw (5) were included in the alignment. The specific priming sites were determined for each allele.
Multiplex PCR development.
To develop a multiplex PCR for the H:i, H:r, H:e,h, H:l,v, H:b, H:z10, and H:d first-phase flagellar antigens, 11 primers were designed with particular concern for the genetic stability of the target sequence, the presence of sequences specific for consensus antigen-specific sites, and the lengths of the amplicons. Primers Forward-G and Reverse-G were designed in order to detect antigens belonging to the G-complex alleles (g,m, g,m,s, g,m,t, g,p, g,p,s, g,m,q, g,q, g,m,p, g,m,p,s, g,m,s,t, g,s,t, m,t, m,p,t,u, f,g,s, f,g,m,t, g,z51, and g,z63). Sdf primer pairs (1) were added in order to detect serotype Enteritidis (9,12:g,m:−). The primers used are shown in Table 3.
TABLE 3.
Primers used for multiplex PCR identification of the most frequent Salmonella first-phase flagellar antigens
| Primer | Sequence (5′-3′) | 5′ position in fliC genea | Reference or source |
|---|---|---|---|
| Sense-60 | GCAGATCAACTCTCAGACCCTGGG | 481-504 | 17 |
| Antisense-i | ATAGCCATCTTTACCAGTTCC | 714-734 | This work |
| Antisense-z10 | CGTCGCAGCTTCTGCAACC | 911-929 | This work |
| Antisense-b | CGCACCAGTCYWACCTAAGGCGG | 628-650 | This work |
| Antisense-eh | AACGAAAGCGTAGCAGACAAG | 658-678 | This work |
| Antisense-lv | CCTGTCACTTTCGTGGTTAT | 790-809 | This work |
| Antisense-r | AAGTGACTTTTCCATCGGCTG | 741-761 | This work |
| Forward-d | CCCGAAAGAAACTGCTGTAACCG | 539-561 | This work |
| Reverse-d | TGGATATCAGTATTGCTCTGGGC | 625-647 | This work |
| Forward-G | GTGATCTGAAATCCAGCTTCAAG | 547-569 | This work |
| Reverse-G | AAGTTTCGCACTCTCGTTTTTGG | 1055-1078 | This work |
| Forward-Sdf-l | TGTGTTTTATCTGATGCAAGAGG | — | 1 |
| Reverse-Sdf-l | CGTTCTTCTGGTACTTACGATGAC | — | 1 |
Dashes indicate that primers do not target the fliC gene.
The multiplex PCR was performed with Ready-To-Go PCR beads (Amersham Biosciences). According to the instructions of the manufacturer, DNA amplifications were performed in a reaction volume of 25 μl. Each reaction mixture contained 1.5 mM MgCl2 and 5 pmol each of primers P60, Reverse-r, Reverse-i, Reverse-l,v, Forward-G, Reverse-G, and Reverse-z10 and the Sdf primer pairs (1). Also, 7 pmol each of primers Reverse-eh, Reverse-b, Forward-d, and Reverse-d was included. To improve the efficiency of the multiplex reaction and to reduce cross-reactivity, the buffer concentration was raised to 1.6× (7). Three microliters of a briefly centrifuged, boiled strain suspension was used as the template. The PCR was carried out under the conditions published previously (5): one initial denaturation cycle at 95°C for 5 min, followed by 30 cycles of 95°C for 40 s, 58°C for 20 s, and 72°C for 20 s and one final extension cycle of 72°C for 7 min. The fragments were separated in 2.5% agarose gels by unidirectional electrophoresis and were visualized by staining with ethidium bromide. Fragment size was determined by comparison with 50- and 100-bp DNA ladders (Amersham Biosciences).
Nucleotide sequence accession numbers.
All ICSs were deposited in GenBank with accession numbers AY429608 to AY429613 and AY434692 to AY434712.
RESULTS
Two consensus sequences for the fliCi 600-bp VIR were determined. These sequences differed at 2 nucleotides, which gave only one change in the amino acid sequences (Thr-195ACG→Ala-195GCG).
In the case of fliCr, two consensus sequences were also determined. Three polymorphic sites were found within the sequences, but these resulted in synonymous substitutions.
Three consensus sequences for the fliCz10 VIR were determined, ICSz10-1, ICSz10-2, and ICSz10-3, within which there were seven polymorphic sites. Four of the substitutions were nonsynonymous. ICSz10-1 and ICSz10-2 differed at only 1 nucleotide, which gave a change in the amino acid sequence (Ala-204GCT→Thr-204ACT). ICSz10-3 and ICSz10-1 differed at 7 nucleotides and 3 amino acids (Thr-204ACC→Ala-204GCT, Ser-205AGT→Gly-205GGT, and Thr-348ACT→Ser-348AGT). ICSz10-3 and ICSz10-2 differed at 6 nucleotides. These changes were translated into two changes in the amino acid sequence (Ser-205AGT→Gly-205GGT and Thr-348ACT→Ser-348AGT, respectively).
Three consensus sequences for fliCl,v were established: ICSlv-1, ICSlv-2, and ICSlv-3. Only two polymorphic sites were found, but they were not involved in amino acid sequence changes.
Three consensus sequences for fliCe,h were determined: ICSeh-1, ICSeh-2, and ICSeh-3. There were four polymorphic sites among them, and all of the sites were nonsynonymous: Ala-185GCT→Asp-185GAT, Ala-215GCG→Thr-215ACG and Ala-215GCG →Lys-215AAG, and Thr-277ACA→Lys-277AAA, respectively.
Three consensus sequences for fliCd were established: ICSd-1, ICSd-2, and ICSd-3. There were five polymorphic sites. One of the sites (Lys-262AAA→Thr-262ACA) was nonsynonymous.
Finally, two consensus sequences were determined for fliCb. There were 27 polymorphic sites. Sixteen of them gave 11 changes in the amino acid sequence: Ala-201GCA→Thr-201ACA, Asp-204GAC→Thr-204ACC, Ala-207GCA→Glu-207GAA, Ala-210GCC→Thr-210ACC, Thr-214ACG→Leu-214TTG, Ser-218TCG→Thr-218ACG, Ala-261GCA→Ser-261TCA, Val-274GTT→Ile-274ATT, Thr-281ACG→Ala-281GCA, and Ala-357GCA→Thr-357ACA. Furthermore, ICSb-2 (corresponding to serotype Wien) had a GCT insertion (Ala-239).
G-complex analysis.
The complete gene sequence was used to study the G complex. We used only one sequence for the fliCg,m,p, fliCg,m,t, and fliCm,p,t,u alleles. The sequences obtained for the fliCg,m, fliCg,m,s, fliCg,p, fliCg,p,u, fliCg,q, and fliCm,t alleles were identical to those published previously (10).
Two consensus sequences were found for fliCg,m,s,t. Both sequences were different at 15 nucleotides. Three of them gave changes in the amino acid sequence: Ala-254GCT→Thr-254ACT, Ala-314GCC→Gly-314GGC, and Val-383GTC→Ile-383ATC.
The fliCg,p,s sequence was different from the previously published sequence (10) for the same serotype (serotype Naestved) by one nonsynonymous nucleotide change (Asn-348AAC→Ser-348AGC).
Two consensus sequences were determined for fliCg,s,t. They were different at 7 nucleotides, although the amino acid sequences were identical. The sequence obtained for serotype Senftenberg was identical to that published previously (10).
The sequences obtained for fliCg,t were different in the two serotypes studied: Agodi and Merseyside. There were 44 polymorphic sites. Eight of them gave changes in the amino acid sequence: Asn-246AAC→Asp-246GAT, Asp-258GAT→Ala-258GCT, Ser-265AGT→Gly-265GGT, Asn-319AAT→Asp-319GAT, Asp-322GAT→Ala-322GCT, Ala-380GCG→Thr-380ACG, Val-384GTC→Ile-384ATC, and Ser-421TCA→Ala-421GCA. The sequence of fliCg,t obtained from serotype Agodi was different from the previously published fliCg,t sequence (10) from serotype Budapest at 9 nucleotides, which resulted in one change in the amino acid sequence Ser-265AGT→Gly-265GGT. Finally, the sequence of fliCg,t obtained for serotype Merseyside was different from the previously published sequence (10) at 40 nucleotides, which gave six different amino acids: Asn-246AAC→Asp-246GAT, Asp-258GAT→Ala-258GCT, Asp-322GAT→Ala-322GCT, Arg-380CGC→Thr-380ACG, Val-384GTC→Ile-384ATC, and Ser-421TCA→Ala-421GCA.
Position numbers are equivalent to those previously reported for the fliC flagellin with antigen i (11).
Primer design.
By comparison of both the DNA sequences and the deduced amino acid sequences, we chose the common forward primer sense-60 for antigens H:i, H:r, H:z10, H:l,v, H:e,h, and H:b and six reverse primers specific for each of these alleles.
In the case of the E complex, the differences were diverse enough that we were able to design a reverse primer specific for the e,h antigen.
The alleles belonging to the L complex showed high degrees of similarity when the sequences within the complex were compared. There were only five nonsynonymous sites in the sequence. However, a reverse primer specific for l,v could be designed.
The sequences of the antigens belonging to the G complex showed high degrees of homology. Because of this homology, it was not possible to design a primer specific for the m antigen. A highly conserved region was used to design a universal forward primer (primer Forward-G) and a universal reverse primer (primer Reverse-G). In order to identify Salmonella serotype Enteritidis (9,12:g,m:−), the serotype reported most frequently in Spain, Sdf primer pairs, previously described by Agron et al. (1) to be specific for this serotype, were added to our multiplex PCR mixture to distinguish Salmonella serotype Enteritidis from other serotypes with H:6 in fliC.
The fliCd sequences were diverse enough with respect to the sequences of the other alleles sequenced and did not match the sense-60 sequence. Consequently, we have designed fliCd-specific forward and reverse primers.
Multiplex PCR sensitivity and specificity.
Each of the reverse primers was tested individually and in combination with its specific forward primer by using a panel of known serotypes to ensure that a PCR product of the expected size was produced and that no additional or nonspecific products were generated. Once the specificity was determined, the PCR conditions, buffers, and primer concentrations were optimized (7) to establish conditions in which the primers could be combined into a single PCR mixture (see Material and Methods).
This new Salmonella phase-1 multiplex PCR developed with the newly designed and previously published (1, 3, 17) primers (Table 3) was evaluated with 161 S. enterica subsp. enterica isolates expressing one of the following flagellar antigens: H:G, H:z10, H:l,v, H:r, H:i, H:e,h, H:b, and H:d. These reactions gave unique and specific amplicons of 500, 400, 300, 275, 250, 200, 150, and 100 bp, respectively.
Two amplicons were identified for Salmonella serotype Enteritidis: one of 500 bp that corresponds to the G complex and one of 333 bp that is specific for this serotype (Fig. 1). No cross-reactions with any of the 12 strains expressing the second-phase H:l,w flagellar antigen or 22 strains expressing the first-phase H:l,v flagellar antigen were found. In addition, no cross-reactivity with 41 Salmonella strains expressing second-phase H:e,n,x and H:e,n,z15 flagellar antigens or 15 Salmonella strains expressing first-phase H:e,h flagellar antigens was found. Finally, no amplification of flagellar antigen alleles was seen from any of the strains in the negative control panel. The correlation of the PCR assay results with the results of traditional serotyping was 100%.
FIG. 1.
Multiplex PCR amplification of Salmonella first-phase flagellar antigens. Lane 1, 100-bp ladder (Amersham Biosciences); lane 2, serotype Derby (H:f,g); lane 3, serotype Enteritidis (H:g,m); lane 4, serotype Mikawasima (H:y); lane 5, serotype Hadar (H:z10); lane 6, serotype Brandenburg (H:l,v); lane 7, serotype Infantis (H:r); lane 8, serotype Typhimurium (H:i); lane 9, serotype Anatum (H:e,h); lane 10, serotype Ohio (H:b); lane 11, serotype Grumpensis (H:d); lane 12, 50-bp ladder (Amersham Biosciences).
DISCUSSION
Traditional serotyping of S. enterica has identified 2,501 unique serotypes based on antigenic surface structures of the lipopolysaccharide (O antigen), the phase-1 flagellar protein (H1), and the phase-2 flagellar protein (H2) (15). In reality, though, each year only a limited number of serotypes are reported to LNRSS. For example, the 10 most common serotypes constitute 89.8% of the clinical isolates reported to LNRSS. Salmonella serotypes Typhimurium and Enteritidis alone represent 73.9% of the clinical isolates (16).
However, serotyping by the traditional K-W scheme is time-consuming, requires well-trained technicians, and uses large amounts of high-quality sera. For these reasons, the use of DNA methods, such as the multiplex PCR for H-antigen identification described here, is an attractive alternative to the more traditional techniques.
During the past decade, many studies have demonstrated the extremely high capacity of PCR to detect specific genes of interest. These studies have shown that PCR can be a powerful tool in clinical microbiology. Echeita and colleagues (5, 6) showed that a multiplex PCR method which identifies the specific second-phase flagellar antigens found in the most common Salmonella serotypes reported in Spain could be developed. We illustrate here that the use of a similar method to identify the first-phase flagellar antigens and the use of a PCR approach to O-antigen typing could identify the major Salmonella serotypes, leaving conventional serotyping to be used for identification of the less common serotypes. The use of these methods would dramatically reduce the reliance on the traditional time-consuming methods and reduce the amount of sera needed for Salmonella identification.
The analysis of first-phase alleles encoding H:i, H:r, H:l,v, H:e,h, H:b, H:z10, and H:d antigens showed the high degree of variability among phase-1 antigens. fliC VIR sequences were variable enough to allow the design of primers specific for each antigen. The presence of more than one primer pair in a multiplex PCR increases the possibility of obtaining spurious amplification products, primarily due to primer-dimer formation (2). For this reason, previously designed primer sense-60 (17) was chosen as a common forward primer specific for alleles fliCi, fliCr, fliCz10, fliCe,h, fliCb, fliCd, and fliCl,v. The remaining primers were designed for genetic stability of the target sequence, specificity for consensus antigen-specific sites, and the lengths of the amplicons.
For detection of the L and E complexes, previously published sequences of fljBe,n,x, fljBe,n,z15, fljBl,w, and fliCl,w (5) were also included in order to design primers specific for the l,v and e,h antigens. The previously reported high degrees of homology within these complexes led us to include these sequences in our alignment in order to detect and eliminate future nonspecific reactions.
E-complex sequence analysis demonstrated that first-phase-antigen amino acid sequences (represented by fliCe,h) and second-phase-antigen amino acid sequences (represented by fljBe,n,x and fljBe,n,z15) were very different. Second-phase allelic differences implicated only three amino acid changes: Ala224 and Ser237, which are probably implicated in the e,n,x epitope conformation, while Gly237 is probably at least partially responsible for the e,n,z15 epitope conformation.
The L-complex sequences were more homogeneous, with only five polymorphic sites among their amino acid sequences. All amino acid substitutions had the same hydrophilic properties. Thr264 and Thr269 are probably implicated in l,v epitope specificity; similarly, Ala197, Thr202, and Ala206 are responsible for l,w antigen specificity.
A high degree of homology existed among the H:G complex (g,m, g,m,s, g,m,t, g,p, g,p,s, g,m,q, g,q, g,m,p, g,m,p,s, g,m,s,t, g,s,t, m,t, m,p,t,u, f,g,s, f,g,m,t, g,z51, and g,z63) VIR sequences, and the sequences diverged from all other gene sequences studied. The fliCg,m allele was found to be so highly conserved that we have been unable to find a region that was not shared with at least one of the other alleles belonging to the H:G complex. Consequently, we designed specific forward and reverse primers that generate an amplicon of 500 bp common to all G-complex antigens. We could not distinguish Salmonella serotype Enteritidis (9,12:g,m:−) phase-1 alleles from other G-complex alleles by this approach. Because this serotype is the most frequent cause of salmonellosis in Spain, as well as in most countries, we believed that it was important to design a multiplex PCR that could also distinguish this serotype from the other G-complex serotypes. For this reason, the Sdf primer pairs that Agron et al. (1) described as being specific for serotype Enteritidis were included in the multiplex PCR.
We report here on a set of both known and novel primers, comprising four forward primers and nine reverse primers, that are used together in a unique multiplex reaction. Using this single PCR assay, we can now rapidly identify the most common first-phase flagellar antigens of S. enterica isolates found in Spain. As our results showed, the method described here is a specific, fast, and cost-effective method that can be applied in a clinical microbiology laboratory for the serotyping of the first-phase flagellar antigens commonly expressed by Salmonella strains. The PCR described in this work, together with the multiplex PCR specific for the second-phase flagellar antigen reported by Echeita et al. (5), could identify the first- and second-phase flagellar antigens of up to 80% of the Salmonella strains isolated in Spain.
REFERENCES
- 1.Agron, P. G., R. L. Walker, H. Kinde, S. J. Sawyer, D. C. Hayes, J. Wollard, and G. L. Andersen. 2001. Identification by subtractive hybridization of sequences specific for Salmonella enterica serovar Enteritidis. Appl. Environ. Microbiol. 67:4984-4991. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Brownie, J., S. Shawcross, J. Theaker, D. Whitcombe, R. Ferrie, C. Newton, and S. Little. 1997. The elimination of primer-dimer accumulation in PCR. Nucleic Acids Res. 25:3235-3241. [Online.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Burnens, A. P., J. Stanley, R. Sack, P. Hunziker, I. Brodard, and J. Nicolet. 1997. The flagellin N-methylase gene fliB and an adjacent serovar-specific IS200 element in Salmonella typhimurium. Microbiology 143(Pt 5):1539-1547. [DOI] [PubMed] [Google Scholar]
- 4.Burnens, A. P., J. Stanley, I. Sechter, and J. Nicolet. 1996. Evolutionary origin of a monophasic Salmonella serovar, 9,12:l,v:−, revealed by IS200 profiles and restriction fragment polymorphisms of the fljB gene. J. Clin. Microbiol. 34:1641-1645. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Echeita, M. A., S. Herrera, J. Garaizar, and M. A. Usera. 2002. Multiplex PCR-based detection and identification of the most common Salmonella second-phase flagellar antigens. Res. Microbiol. 153:107-113. [DOI] [PubMed] [Google Scholar]
- 6.Echeita, M. A., and M. A. Usera. 1998. Rapid identification of Salmonella spp. phase 2 antigens of the H1 antigenic complex using “multiplex PCR.” Res. Microbiol. 149:757-761. [DOI] [PubMed] [Google Scholar]
- 7.Henegariu, O., N. A. Heerema, S. R. Dlouhy, G. H. Vance, and P. H. Vogt. 1997. Multiplex PCR: critical parameters and step-by-step protocol. BioTechniques 23:504-511. [DOI] [PubMed] [Google Scholar]
- 8.Kilger, G., and P. A. Grimont. 1993. Differentiation of Salmonella phase 1 flagellar antigen types by restriction of the amplified fliC gene. J. Clin. Microbiol. 31:1108-1110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lino, T. 1977. Genetics of structure and function of bacterial flagella. Annu. Rev. Genet. 11:161-182. [DOI] [PubMed] [Google Scholar]
- 10.Masten, B. J., and T. M. Joys. 1993. Molecular analyses of the Salmonella g. . .flagellar antigen complex. J. Bacteriol. 175:5359-5365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.McClelland, M., K. E. Sanderson, J. Spieth, S. W Clifton., P. Latreille, L. Courtney, S. Porwollik, J. Ali, M. Dante, F. Du, S. Hou, D. Layman, S. Leonard, C. Nguyen, K. Scott, A. Holmes, N. Grewal, E. Mulvaney, E. Ryan, H. Sun, L. Florea, W. Miller, T. Stoneking, M. Nhan, R. Waterston, and R. K. Wilson. 2001. Complete genome sequence of Salmonella enterica serovar Typhimurium LT2. Nature 413:852-856. [DOI] [PubMed] [Google Scholar]
- 12.Nair, S., T. K. Lin, T. Pang, and M. Altwegg. 2002. Characterization of Salmonella serovars by PCR-single-strand conformation polymorphism analysis. J. Clin. Microbiol. 40:2346-2351. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Newton, S. M., R. D. Wasley, A. Wilson, L. T. Rosenberg, J. F. Miller, and B. A. Stocker. 1991. Segment IV of a Salmonella flagellin gene specifies flagellar antigen epitopes. Mol. Microbiol. 5:419-425. [DOI] [PubMed] [Google Scholar]
- 14.Ohl, M. E., and S. I. Miller. 2001. Salmonella: a model for bacterial pathogenesis. Annu. Rev. Med. 52:259-274. [DOI] [PubMed] [Google Scholar]
- 15.Popoff, M. Y. 2001. Antigenic formulas of the Salmonella serovars, 8th ed. Institute Pasteur, Paris, France.
- 16.Usera, M. A., A. Aladueña, R. Díez, M. de la Fuente, R. Gutierrez, P. Cerdán, M. Arroyo, R. González, and A. Echeita. 2001. Análisis de las cepas de Salmonella spp. aisladas de muestras clínicas de origen humano en España del 2000 (I). Bol. Epidemiol. Semanal. 9:221-224. [Google Scholar]
- 17.Vanegas, R. A., and T. M. Joys. 1995. Molecular analyses of the phase-2 antigen complex 1,2. . of Salmonella spp. J. Bacteriol. 177:3863-3864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wei, L. N., and T. M. Joys. 1985. Covalent structure of three phase-1 flagellar filament proteins of Salmonella. J. Mol. Biol. 186:791-803. [DOI] [PubMed] [Google Scholar]
- 19.Zieg, J., M. Silverman, M. Hilmen, and M. Simon. 1977. Recombinational switch for gene expression. Science 196:170-172. [DOI] [PubMed] [Google Scholar]