Molecular identification in monophasic and nonmotile variants of Salmonella enterica serovar Typhimurium

Variant strains of Salmonella enterica serovar Typhimurium, lacking one or both flagellar phases have been widely reported. The monophasic S.1,4,[5],12:i:- variant has emerged worldwide in the past few years and has become one of the most frequently encountered in many countries. In contrast, monophasic S.1,4,[5],12:-:1,2 and nonmotile S.1,4,[5],12:-:- strains are rarely described. This study investigated seven molecular markers to identify and delineate monophasic S.1,4,[5],12:i:- (n = 90), S.1,4,[5],12:-:1,2 (n = 25), nonmotile S.1,4,[5],12:-:- (n = 17) strains, and some serovar Typhimurium strains (n = 124) collected through the French Salmonella network between 2001 and 2010. Three markers were commonly detected in serovar Typhimurium and in all variant strains: STM2757, mdh and fliA-B. Monophasic S.1,4,[5],12:i:- were genotypically confirmed by the absence of the fljB, fljA, and hin genes. Nevertheless, 13 (14.5%) of them were positive for these last three genes, revealing monophasic strains named “inconsistent” as previously described. All nonmotile 1,4,[5],12:-:- strains had the fliC, fljA, fljB, and hin genes and the fliC gene was detected in 88% of monophasic S.1,4,[5],12:-:1,2 strains. The combination of the seven markers detection enables to recognize eight different genotypes within the S.1,4,[5],12:i:- collection, among which the Spanish and the U.S. clones previously described could be distinguished and assigned to a genotype. Based on this molecular approach, 71% of the French S.1,4,[5],12:i:- collection belonged to the Spanish clone, whereas only 2% were assigned to the U.S. clone. This study highlights the usefulness of these molecular markers and genotypes for identifying lineages, especially among the epidemiologically important monophasic S.1,4,[5],12:i:- variant.


Introduction
The main reservoir of Salmonella enterica is the intestinal tracts of various animal species and isolates are recovered from food and animal products such as raw meat, poultry products, and milk-and egg-based products (Swaminathan et al. 2006). This Gram-negative flagellated bacterium is responsible for salmonellosis, a food-borne disease causing enteric illness that can sometimes lead to hospitalization. In 2009, approximately 110,000 human cases of salmonellosis have been reported in Europe (EFSA and ECDC 2011). Within S. enterica species, more than 2600 serovars have been reported and the most frequently isolated in human cases worldwide are Enteritidis and Typhimurium.
Together, they cause around 80% of confirmed cases of salmonellosis in Europe (EFSA and ECDC 2011). In France, 38.9% of human isolates collected by the National Reference Centre for Salmonella in 2009 (NRC-Salm) and 13% of food and in veterinary isolates collected by the French Salmonella Network were from serovar Typhimurium. Salmonella serovars are defined with an antigenic formula based on the presence of somatic and flagellar antigens according to the White-Kauffmann-Le Minor scheme (Grimont and Weill 2007). In most S. enterica subsp. enterica serovars, the antigenic formula is composed of two flagellar phases. The first flagellar phase is encoded by the fliC gene and the second one is encoded by the fljB gene. Expression of fliC and fljB is regulated through a mechanism called "phase variation" mediated by a DNA invertase, Hin, involved in the reversible inversion of the H segment, a DNA segment containing the promoter for the fljB gene (Fig. 1). This inversion occurs by site-specific recombination between two inverted repeat sequences flanking the H segment, hixL and hixR ). The fljA gene, which encodes a negative regulator of fliC expression, is located downstream of fljB. When the H segment is in the "on" orientation, both fljB and fljA are transcribed. Only the second flagellar phase is expressed because fliC expression is repressed by FljA. However, when the H segment is in the "off" orientation, neither fljB or fljA are transcribed, so the fliC gene is expressed and only the first flagellar phase is phenotypically detected (Yamamoto and Kutsukake 2006). Moreover, a complete genomic sequence study revealed that the serovar Typhimurium strain LT2 harbored another DNA invertase gene named fin gene, and located to a 2.7-kb invertible DNA fragment within a resident prophage, Fels-2 .
Monophasic variant lacking second flagellar phase S.1,4, [5],12:i:-emerged worldwide a few years ago ) and have since become one of the most frequently isolated serovars in many countries. In France, this monophasic variant was the third most isolated serovar in humans in 2009, behind serovars Typhimurium and Enteritidis (F. X. Weill, S. Le Hello, Annual report from the NRC-Salm). In human cases as well as in food and animal products, isolation rates have increased 10-fold from 2005 to 2010 (NRC-Salm and ANSES Salmonella Network data). Furthermore, two other S.Typhimurium-like variants that are not frequently encountered have been described, one lacking the first flagellar phase S.1,4, [5],12:-:1,2 and the nonmotile variant, S.1,4,[5],12:-:-.

482
Therefore, the development of rapid methods adapted to the detection of such monophasic or nonmotile variants strains is critical for surveillance and identification of the different lineages is needed for food safety risk assessments and clinical microbiological investigations. In this study, a total of 256 strains of serovar Typhimurium, monophasic and nonmotile variant of S.Typhimurium-like isolated from food and animal origins were investigated. Seven molecular markers including those recommended by EFSA were used to improve the molecular characterization of the different variants and to determine the genetic profile of these strains.

PCR detection
Before DNA extraction, strains were cultured on Drigalski agar (BioMérieux, Marcy-l'Etoile, France). After overnight incubation at 37°C, DNA from a single pure colony was extracted using Instagene Matrix ® (Bio-Rad Laboratories, Marnes-la-Coquette, France) according to the manufacturer's recommendations.
The fliA-B intergenic region and the fliC and fljB genes were detected with conventional PCR protocols based on EFSA recommendations (EFSA 2010) using the following primer pairs: FFLIB and RFLIA, Sense-59 and Antisense-83, and Sense-60 and Antisense-I, respectively. Each conventional PCR was performed in simplex using an Applied Biosystems Veriti cycler (Applied Biosystems, Carlsbad, California). PCR detection of the fliA-B intergenic region was performed in 19 PCR buffer, 3.5 mmol/L MgCl 2 , 0.3 mmol/L dNTPs, 0.6 lmol/L each primer (FFLIB and RFLIA) (EFSA 2010), and 1 U of FastStart Polymerase (Roche, Meylan, France). The cycling profile for this PCR involved denaturation at 94°C for 5 min, followed by 30 cycles of 94°C for 30 sec, 64°C for 30 sec, and 72°C for 90 sec, and a final step of 72°C for 10 min. PCR amplification of the fljB gene encoding a 1389-bp product linked to the phase-2 antigen determinant was performed in 19 PCR buffer, 2.5 mmol/L MgCl 2 , 0.3 mmol/L dNTPs, 1 lmol/L each primer (Sense-59 and Antisense-83) (EFSA 2010), and 1 U of FastStart Polymerase. The cycling profile for this PCR involved denaturation at 94°C for 5 min, followed by 30 cycles of 94°C for 30 sec, 62°C for 30 sec, and 72°C for 30 sec, and a final step of 72°C for 7 min. Amplification of the fliC gene encoding a 550-bp product linked to the phase-1 antigen determinant was performed in 19 PCR buffer, 1.5 mmol/L MgCl 2 , 0.2 mmol/L dNTPs, 0.5 lmol/L each primer (Sense-60 and Antisensei) (Herrera- Leon et al. 2004;Levy et al. 2008), and 1 U of FastStart Polymerase. The cycling profile for this PCR involved denaturation at 94°C for 5 min, followed by 30 cycles of 94°C for 40 sec, 58°C for 20 sec, and 72°C for 20 sec, and a final step of 72°C for 7 min.
Moreover, real-time PCR reactions were designed to target the five following markers, the ttrC gene as Salmonella genus marker (Malorny et al. 2004), two serovar Typhimurium-associated genes: STM2757 and mdh (malic acid dehydrogenase gene) (Amavisit et al. 2005;Soyer et al. 2009;Hopkins et al. 2010), the fljA gene as fliC repressor and the hin gene as DNA invertase of the H segment ). PCR reactions were simultaneously performed using the Fluidgm Biomark TM realtime PCR system (Fluidigm ® , San Francisco, California) with the 96.96 dynamic arrays TM . This array system has the capacity to screen 96 DNA templates against 96 genetic markers in the same run. Amplifications were performed according to the manufacturer's recommendations using EvaGreen ® DNA-binding dye (Biotium Inc, Hayward, California) and were followed by a melting curve analysis. The BioMark TM real-time PCR system was used with the thermal profile of 95°C for 10 min, followed by 35 cycles of 95°C for 15 sec, and 60°C for 1 min. Results were analyzed using the Fluidigm software packages "Fluidigm Melting Curve Analysis" (Version 3.0.2) and "Fluidigm Real-Time PCR Analysis" (Version 3.0.2).

Sequences
Primer design and names were based on the nucleotide sequence of S. enterica enterica Typhimurium LT2 referenced in GenBank with the accession number AE006468. Primer sequences are listed in Table 1.

Detection of markers common to serovar Typhimurium and to the variant strains
The 256 investigated strains as well as the LT2 strain, the Farsta, Gloucester, and Lagos were all positive for ttrC, which is a Salmonella genus marker. The three different markers (fliA-B, STM2757, and mdh) usually associated with serovar Typhimurium were explored on the same collection. As expected, the three strains of serovars Farsta, Gloucester, or Lagos were negative for these three markers (data not shown). The 124 strains of serovar Typhimurium all were positive for markers STM2757 and mdh. The 1-kb fliA-B intergenic region was also detected in all serovar Typhimurium strains except one. This latter fliA-B-negative strain was further serologically confirmed.  12:-:1,2 strains, respectively. Six of the seven STM2757negative S.4,12:i:strains had been isolated from pork products, and the single remaining strain had been recovered from poultry. The two S.4,5,12:-:1,2 STM2757negative strains had been isolated from effluents. The absence of detection of this marker was investigated using primers flanking STM2757 (STM2757-seqF and STM2757-seqR). The targeted locus cannot be amplified, revealing that the entire STM2757 ORF and its vicinity were absent (data not shown). Finally, all three Typhimurium-associated markers, fliA-B, STM2757, and mdh, were detected in the nonmotile strains (n = 17) ( Table 2).
Detection of the genes involved in flagellar structure (fliC and fljB) and in their expression regulation (fljA and hin) All serovar Typhimurium strains including the reference LT2 strain, but one, carried both fliC and fljB genes encoding the first and second flagellar phases. Indeed, only a single Typhimurium strain did not display the fljB marker. This fljB-negative Typhimurium strain had been isolated from a ready-to-eat meal and was positive for the presence of the fljA and hin markers, targeting genes flanking fljB (Fig. 1). This strain was serologically confirmed as serovar Typhimurium, indicating that the second flagellar phase antigen was detected at the bacterial surface. As expected, all S.1,4,[5],12:i:strains carried the fliC gene involved in the structure of the first flagellar phase, and 81% (n = 73) of them were PCR negative for the gene encoding the second flagellar phase, fljB (Table 2) Table 2). In contrast, fljA and hin markers were generally not detected in S.1,4,[5],12:i:strains. These results are fully coherent because fljA is located in the operon of the second flagellar phase and encodes the repressor of the expression of the first flagellar phase. These results are in total agreement with other previous studies (Echeita et al. 2001;Tavechio et al. 2004;Soyer et al. 2009;Hauser et al. 2010;Laorden et al. 2010;Trupschuch et al. 2010;Ido et al. 2011).
Furthermore, strains of serovars Farsta, Gloucester, and Lagos showed positive amplification of the fliC and fljB markers, which corroborates their phenotypic biphasic structure. The two markers fljA and hin were also detected in these last serovars (data not shown).
Moreover, this study identified three other patterns that have never been described before. The P4 pattern (STM2757+, mdh+, fliAB+, fliC+, fljB+, fljAÀ, and hinÀ), highly similar to those described in Spanish clone strains was observed in four strains from pork (n = 2) and chicken (n = 2) sources. The only difference lies in the presence of the fljB marker. Strains carrying this pattern should be subjected to numerous recombination events leading to the conservation of the fljB gene and deletion of the fljA and hin genes that flank it (Fig. 1). However, the absence of the fliC repressor gene (fljA) allows constitutive expression of the first flagellar phase. The absence of the hin invertase gene does not allow the switch between the expression of the first and second flagellar genes. Therefore, with regard to the serological result, the H segment may be locked in the "off" orientation ( Fig. 1) (Yamamoto and Kutsukake 2006;Switt et al. 2009). Four other strains from pork products (n = 3) and poultry (n = 1) were assigned to P3 pattern (STM2757+, mdh+, fliAB+, fliC+, fljBÀ, fljA+, and hin+) related to the Typhimurium P1 pattern, differing only by the absence of the fljB gene. Finally, three monophasic S.1,4,[5],12:i:strains recovered from pigs (n = 1), poultry (n = 1), and pork products (n = 1) exhibited the P6 pattern (STM2757+, mdh+, fliAB+, fliC+, fljBÀ, fljA+, and hinÀ).

Discussion
This study aimed to investigate three different molecular markers already described for the identification of serovar Typhimurium.
One of these markers, the fliA-B intergenic region proposed by the EFSA working group (EFSA 2010) enables the identification of all serovar Typhimurium strains except one. This latter fliA-B-negative strain was further serologically confirmed to be serovar Typhimurium. The fliA gene encodes the flagellum-specific sigma factor r 28 of Salmonella Typhimurium (Ikebe et al. 1999) and fliB gene encodes a flagellin methylase (Frye et al. 2006). So, FliA regulates the expression of genes involved in the structure and the assembly of both flagella, whereas the flagellin methylation carried out by FliB is not involved in mobility. This methylation should be required for Salmonella virulence and not for flagellin function. The only marker detected in all serovar Typhimurium strains including the monophasic and nonmotile variants was mdh, whereas the STM2757 and fliA-B markers amplified in most but not all of these strains at a rate of 96.5% and 99.6%, respectively. Furthermore, after the investigation of the mdh marker on a collection of 937 strains belonging to a large panel of more than 230 different serovars, no cross-reaction was detected except for a single isolate of serovars Kibusi and New Mexico; both these serovars are very rarely reported worldwide.
Then, the implicated genes in the flagellar structure and the expression were investigated. While Typhimurium strains should possess all tested flagellar genes, a fljB-negative Typhimurium strain was identified. This strain was positive for the presence of the fljA and hin markers, targeting genes flanking fljB (Fig. 1) and has been serologically confirmed as serovar Typhimurium, indicating that the second flagellar-phase antigen was well detected at the bacterial surface. On the other hand, some monophasic S.1,4,[5],12:i:strains were fljB positive. Such strains already reported by Hopkins et al., Soyer et al. and Bugarel et al. are considered to be "inconsistent" variants (Soyer et al. 2009;Hopkins et al. 2010;Bugarel et al. 2012) as they are characterized by an inconsistent serological detection of the second flagellar phase. These findings 486 could suggest that such strains may have deletions or mutation in the fljB gene out of the primer-binding site of the fljB marker that modifies protein structure inhibiting serological detection. Nevertheless, sequencing data is requested to consolidate this hypothesis. Moreover, the invertible promoter controlling fljB and fljA expression should be locked in one position, allowing only the expression of the first flagellar phase as suggested by Hopkins et al. (2010). Another invertase, Fin, has been described in Salmonella Typhimurium LT2 reference strain. The fin invertase gene is located in the Fels-2 prophage at 34.5 kb of the hin-fljBA locus. Hin and Fin invertases share 62% of their amino acids and their hix and fix recombination sites share 81% of their nucleotide sequence ). In the absence of the Hin invertase gene, Fin can carry out the inversion of the H segment ).
In the same way, the divergence between the phenotypic and genotypic characterization of S.1,4,[5],12:-:1,2 strains could be explained by a deletion or a mutation in the fliC gene that does not affect the PCR amplification of the fliC marker, but affects serological detection of the surface antigen. Another explanation could be that the invertible promoter is locked in the gene activation position and thereby promotes fljB gene expression. A dysfunction of the promoter invertase Hin or a mutation in the hix recombination sites may be involved.
The phenotypically nonmotile strains also harbored a Typhimurium-like profile for the flagellar antigenic markers. Even if the four markers of the structure and expression regulation of both flagellar phases were detected, other explanations of such nonmotile function could be linked to the failure of some hook-basal body genes involved in the flagellar assembly.
Finally, combination of the seven molecular markers enables to distinguish 11 patterns, of which some of them could be related to some previously described lineages within the emerging monophasic S.1,4,[5],12:i:variant. Thus, this study highlighted that 71% of the French S.1,4, [5],12:i:-strains assigned to P5 or P8 patterns could derive from the previously described Spanish clone, whereas only 2% of these monophasic strains assigned to P7 and P9 patterns could be attributed to the U.S. clone. Moreover, 13 fljB-positive S.1,4,[5],12:i:strains (14.5%) carried the Typhimurium-associated P1 pattern (Table 2) gathering the "inconsistent" strains according to the Hopkins et al. definition ). This particular group of strains was also found in one of our previous studies (Bugarel et al. 2012). Our investigation revealed for the first time the presence of the hin gene in the "inconsistent" S.1,4,[5],12:i:group of strains. Thus, based on their assignment to pattern P1, such strains could be genotypically considered as Typhimurium-like strains with a dysfunction in gene expression. Therefore, absence of the phase-2 protein is not due to the absence of the encoding gene, but should be related to dysfunction in gene expression, in protein translocation, in flagellar filament assembly (Bugarel et al. 2012).

Conclusion
Besides the identification of two previously described clones (U.S. and Spanish clones) within the French monophasic S.1,4,[5],12:i:variant collection over the past 10 years, we evidenced three new, previously undescribed genotypes among the monophasic variant strains assigned to P3, P4, and P6 patterns. Inconsistent variants of serovar Typhimurium, serologically monophasic but genetically Typhimurium like, were observed in 14.5% (n = 13) of the French monophasic lacking the second flagellar phase. Furthermore, S.1,4,[5],12:-:1,2 and S.1,4,[5],12:-:mainly carry the same P1 pattern as Typhimurium, in 88% and 100% of the investigated isolates, respectively, suggesting that these serologically monophasic or nonmobile strains are genetically Typhimurium-like. To test this hypothesis, fliC, fljA, fljB, and hin genes detected in the monophasic and nonmotile phenotypes variant strains need to be sequenced. Proteomic studies also could provide clues as to the functionality of the proteins involved. Finally, this study highlights the usefulness of the seven described molecular markers to delineate genotypes and identify lineages, especially among the epidemiologically important monophasic S.1,4,[5],12:i:variant.