Salmonella Typhimurium in Iran: Contribution of molecular and IS200 PCR methods in variants detection

Salmonella Typhimurium, a zoonotic pathogen, is regarded as a major health and economic concern worldwide. Recently, monophasic variants of this serovar have been significantly associated with human gastroenteritis outbreaks globally, making its accurate identification essential for epidemiological and control purposes. We have identified and analyzed 150 S. Typhimurium from 884 Salmonella genus isolated from humans, domestic animals, poultry, food items and abattoirs origins. The Salmonella isolates were obtained from Iranian National Veterinary Reference Laboratories of 9 provinces during 2007–2016, and from five hospitals in Tehran in 2015. The isolates were evaluated biochemically, serologically, and by PCR amplification of invA, mdh, STM4492, fliC, fljA, fljB, hin genes, IS200 and DT104. invA and mdh genes were used to confirm the S. Typhimurium serotype, fliC and fljB genes for determination of monophasic variants and amplification of IS200 to discriminate the monophasic variants from the closely related serotypes. We identified 78.6% (118/150) as classical S. Typhimurium (fliC, fljB and IS200 positive), 12.6% (19/150) were IS200 negative from all isolates. DT104 is another marker for S.Typhimurium serovar typing. Contrary to EFSA guidelines 20.6% (19/29) of human isolates that lacked IS200 insertion sequence, were confirmed as S.Typhimurium. Compared to the North American/European isolates the low prevalence of fljB negative 6% (9/150) and the high abundance of fliC negative 23.3% (35/150) isolates also were indicative of a different regional atypical population. Studies have shown that the prevalence of monophasic (fljB-) S. Typhimurium worldwide is promoted by the Swine industry. Thus, one reason for this high number of different atypical strains could be inhibition of swine breeding system (house hold and industry) in Iran. These results demonstrate a need for a modified identifying protocol to overcome the regional differences.

Introduction investigation of food-borne diseases [15]. However, this method is costly, time consuming and labor-intensive, while the procedure for detection of phase variation is technically demanding. Therefore, various DNA-based serotyping methods have been developed as alternative/complementary procedures to conventional sero-agglutination [10,16]. Furthermore, the emergence and prevalence of Salmonella isolates lacking expression of phase 1 and/or phase 2 flagellar antigens have made the use of molecular methods essential [10]. PCR-based methods for identification and serotyping of S. Typhimurium are easy to perform, relatively inexpensive and reproducible [15,17]. For serovar identification, amplification of Salmonella-specific invA, and for phase 1 and 2 antigens, fliC, fljA, fljB, fliB-fliA and hin genes have been proposed [18,19]. Although, recently STM 4492 coding for a putative cytoplasmic protein, and malic acid dehydrogenase (mdh) genes have been suggested as more suitable targets for Typhimurium serovar identification [20,21].
Salmonella Typhimurium DT (definitive type) 104 carries multiple drug resistance genes on an integrated DNA segment in the genomic island 1. Salmonella Typhimurium receptor for phage type 104 has been detected since 1994 and reported worldwide due to increasing consumptions of related antibiotics. The receptors are flagella such as FliC and FljB on the bacterial surface [22] and identification of this phage is now one of the diagnostic tools for subtyping S. Typhimurium serovar and increasing antibiotic resistance trait [23].
To discriminate the S.Typhimurium variants from their closely-related serovars that share the same antigenic formula and phase one flagellar antigen, amplification of fliB-fliA intergenic region which carries a serovar specific copy of IS200 insertion element and produces a 1000-bp amplicon instead of a 250-bp product is recommended by EFSA [10,14,19,21,23,24] Bacterial transposons are mobile genetic elements that promote DNA rearrangements and horizontal gene transfer of the host organism. Ellis et al. have shown that IS200 encodes (707bp gene size) transposase mRNA (tnpA). The sRNA coded in mRNA, starts in 5'UTR of the gene, has a trans-acting effect repressing the expression of genes encoded within Salmonella Pathogenicity Island 1 (SPI-1) mRNA such as invF. However, despite its negative impact on the invasion of the bacteria into the non-phagocytic cells, it is recognized as a pathogenic marker [25].
In Iran, no published data are available on isolation of S.Typhimurium variant strains. Therefore, this study is aimed to investigate the variant strains among the isolates collected from humans, domestic animals, poultry, slaughter houses, red meat, chicken paste and nondomesticated birds, using sero-agglutination and PCR methods.

Sample collection
The National Veterinary Reference Laboratories (NVRL), based in different provinces of Iran (S1 Fig), sent 819 Salmonella samples to the Iran Veterinary Organization from 2007-2016. The NVRL network is responsible for supervision and surveillance of farm animals' health and diseases as well as the hygiene of the dependent industries in Iran. The samples had been identified as Salmonella at genus level through routine and random sampling using biochemical tests [26], and had been stored for further studies. Therefore, no live animals were used in our research. The origin of animals related samples were: feces from industrial animals [cattle, sheep, poultry (broiler, layer)]; fecal droppings from garden birds (sparrow, geese, duck, canary, parrot); the unprocessed food portions such as, red meat and chicken; the processed foods such as chicken paste; and water effluent from the slaughterhouses. There were no humans or human samples directly involved in this study. Sixty five Salmonella samples were isolated from outpatients in microbiology laboratories of five different hospitals during 2015. The isolated Salmonella samples were collected by Microbiological Laboratory of the hospital and handed to us with no information or codes regarding the patient's identity.

Ethical approval
Although no human samples or live animals were involved in this study, this proposal has been reviewed and approved by the committee on the ethics of animal experiments of the Pasteur Institute of Iran (permit number: IR.PII.REC. 1394.83).

Cultivation and seroagglutination
All Salmonella isolates were confirmed at the genus level using selective media, biochemical and molecular analysis [10,19]. The isolates were cultivated overnight from transferred media into the Nutrient Broth, and subsequently subcultured on XLD selective media. A single colony from each plate was used for seroagglutination using polyclonal antibody against O antigen for B serogroup antigen (Mast, UK). The positive isolates were tested by polyclonal antibody against O4 antigen using purchased antisera (Mast, UK).

Molecular identification and serotyping
A single colony from all the 884 isolates, irrespective of seroagglutination results, was sub-cultured in Luria Bertani broth (Merck, Germany) and DNA extracted using a PCR template extraction kit (11796828001 Roche, Germany), according to the manufacturer's protocol. Purity and concentration of the extracted DNA was determined by spectrophotometer (Nano-dropX100, USA), and 10-20 ng was used as PCR templates.
Specific primer pairs for invA, mdh, STM4492 and DT104 genes were used for identification and subtyping and 150 Salmonella of the 884 isolates were confirmed as S.Typhimurium ( Table 1). Seroagglutination of phase 1 antigen was performed on the 150 confirmed Salmonella isolates using anti H(i) antisera (Mast, UK) ( Table 1). Phase 2 flagellar antigen was tested only for the S.Typhimurium isolated from humans and STM4492PCR-negative isolates using antisera from Mast, UK.
fliC, fljA, fljB, and hin genes, as well as fliA-fliB intergenic region were amplified to identify the S.Typhimurium serovar and its variants ( Table 2). Amplifications were carried out in 20 μl using 2X master mix (Fermentas, Lithuania) with denaturation at 94˚C for 5 min and 30 cycles of 94˚C for 60 sec, annealing for each primer pair was as specified in the Table 2 for 30 sec and extension at 72˚C for 1 min, followed by a final extension at 72˚C for 10 min. Genomic DNA  from S. Typhimurium (ATCC14028) and S. Enteritidis (ATCC13076) were used as positive and negative controls, respectively. Amplifications were repeated at least twice and each round of PCR contained a tube with no template DNA to monitor the potential contamination in the PCR components.

Results
Out of 819 animals and 65 human Salmonellas isolates, 121 and 29 were positive for O4 (group B) antigen and PCR positive for invA and mdh genes respectively and were identified as S.Typhimurium (Table 1). The STM4492gene was not detected in 6.6% (10/150) of the strains. One of the STM4492 negative strains belonged to human isolates (Table 3), three to red meat (Table 4), one to sheep (Table 5), four to chicken and one to chicken paste isolates ( Table 6). Sero-agglutination tests for phase 1 and 2 flagellar antigens were repeated on these isolates, confirming the original identification and phase 2 serotyping was positive for one human, one red meat and one sheep sample. Presence of flagellar genes was investigated with primers recommended by the EFSA [10], Amplification and expression of phase one flagellar antigen was observed in 79.3% (23/29) Salmonella Typhimurium from patients (Table 3). Of the remaining 6 isolates, 4 contained phase twoflagellar gene (fljB) but serotyping results were negative and two were both genetically and phenotypically negative. The concomitant presence of each fljB and fliC genes and positive seroagglutination occurred in 86.2% (25/29) and 100% (29/29), respectively ( Table 3).
Identification of DT104 subtype of S. Typhimurium showed that 43.3% (65/150) of S. Typhimurium isolates belonged to DT104 subtype. The results from five hospitals has shown that 95.0% (19/20) of positive DT104 isolates were collected from one hospital with three different flagellar genotype variants. The most prevalent clone of human patient isolates are diphasic isolates 58% (17/29) and the prevalence of negative phase one or two was similar in the remaining isolates(20.6% (6/29) Table 3).
The fljB gene was detected in 92.8% (39/42) of isolates from red meat and the PCR product for intergenic region (fliA-fliB) in all of these strains was the expected 1000 bp, except two isolates which produced a 250 bp amplicon and 11.9% (5/42) of the isolates from red meats sold in shops were of DT104subtype (Table 4). But 100% (18/18) of the Salmonellas strains from live cattle and sheep were of DT104 phage type (Table 5).
In Sistan province, 87.5% (7/8) of isolates in 2011 showed 250-bp amplicon for fliA-fliB, which is the highest, and had different genetic pattern for flagellar genes. Out of these isolates one was from red meat (Table 4), five from cattle (Table 5), and one from chicken (Table 6). Table 5 shows that 100% of the isolates from cattle, sheep and slaughter house carried fljB gene and 86.3% (19/22) were positive for DT104, which was the highest among the sample groups.
The most variation of S.Typhimurium isolates in genotype and in distribution belonged to the chicken group. The gene, fljB, was not found in 28.8% (13/45) of the isolates of poultry and chicken paste, but 23% (3/13) of these isolates showed 250bp amplification products for fliA-   Table 6).The majority of the genetically non motile S.Typhimurium[77.7% (7/9)] were found in the poultry group and the remaining two genetically non motile isolates belonged to the red meat (R77 and R80). The expression of phase one flagellar antigen was not observed in 42.2% (19/45) of the isolates; however, there is a discrepancy where one isolate(C3) gave positive agglutination results, even though the gene for 'i' antigen was not detected by PCR ( Table 6). The presence of DT104gene cluster and the lack of IS200 were detected in 42.2% (19/45) and 11.1% (5/45) of the isolates, respectively (Table 6). Slide agglutination on 47 bacteria isolated from chicken samples with antisera for phase one showed negative result in 42.5% (20/47) of the isolates (Table 6) which was the highest among all the categories.
All 12 Salmonella samples in wild birds group, that were isolated in 2009 were positive for both phase 1 and 2 antigen genes by PCR and the phase 1 by slide agglutination. All the isolates had fliA-fliB region of 1000bp (Table 7) The results of testing different markers and genes produced 11 genetic groups of variants. The most common isolated Salmonella variant (group I) was genetically positive for all the tested genes (Table 8). Followed by VII variants (23.3% (35/150) lacking fliC gene and protein) ( Table 8).

Discussion
Effective surveillance and control of pathogens require accurate identification and strain characterization. In this study, 150 out of 884 Salmonella isolates were identified as S.Typhimurium based on conventional and PCR methods, of which 121 were of non-human origin sent to the National Veterinary Laboratory and 29 from human diarrheal cases. S124, S125, S127, S128, S129, S130, S131, S132, S133, S134, S135  In this study mdh gene was amplified in all 150 isolates whereas, STM4492 which had been suggested by Saeki et al. to be able to differentiate S.Typhimurium from 22 other Salmonella serovars failed to produce an amplicon in 6.6% of isolates [21].
All the isolates identified as S.Typhimurium, were tested for the presence of structural (fliC and fljB) and regulatory genes (fljA and hin) by PCR and phase one flagellar antigen (H(i) using seroagglutination [13].The results of serology and PCR for phase one antigen were the same except for 1.3% (2/150) of the isolates differing from Bugarel et al who reported a 12% (3/25) similarity between the results of the two tests (serology and gene absence for phase one) for S 1,4,[5],12:-:1,2 strains [19].
Based on the molecular results, 11 variants were identified. Similar to other studies Variant I was the most prevalent group. Although, monophasic 1,4, [5],12:i:-variant of S.Typhimurium has emerged as an important cause of salmonellosis, and it has been observed in many countries worldwide, this is the first report concerning their identification in Iran. As the prevalence of genotypically negative fljB was 12% (18/150) (variants II,III,IV, X,XI) of which 66.66% (12/18) were positive for IS200 (Table 8), therefore, 8% (12/150) could be monophasic as recommended by EFSA.
In our study, following classical isolate, the monophasic serotypes of fliC was the second most prevalent group followed by fljB. The latter result is not in agreement with that reported by Bugarel et al. which showed that in France the genotypically negative fljB 28.9% (74/256) is the second most prevalent variant following the classical group [19].
Phase switching of flagellar protein expression is a common feature of Salmonella species [13]. In two isolates of variants III and IV that fljA gene is not detectable, the expression of fliC is in on position, while the fljB could still be subject to the phase variation (Table 8). For IV, VI and VIII variants that the hin gene is not present if the promoter for fljB gene is active, the expression of H(1,2) is locked in on or off position.
In this study, four out of 29 isolatedS.Typhimurium1,4, [5],12:i:-were from hospital patients that carried fljB gene, similar to that reported by Bugarel et al. [19] as "inconsistent" variant. The reason for this could be mutation(s) in the fljB gene out of the primer-binding site affecting protein structure and consequently inhibiting the serological detection or deletion of promoter of fljB gene [27]. These isolates were genotypically considered as Typhimurium-like Salmonella Typhimurium in Iran strains with some flaw in gene expression [19].   Table 2) were genotypically fljB and fliC negative, respectively [19,28]. Whereas, in our study using the same primers, the fljB and fliC negative isolates from various origins were 10.6% (16/150) and 27.3% (41/150), respectively. This result could represent a new pattern for prevalence of S.Typhimurium variants in Iran.
One of the parameters of a serotype spread in a community is the animal industry, studies have shown that the first isolates of monophasic fljB negatives were from poultry (1986) and later detected in pig industry (2003) [15]. Moreover, different studies have shown that the majority of the monophasic Salmonella strains (1,4,[5],12:i:-) were isolated from pigs and pork industry in Spain(1999) [28], beef and pork products in China(2015) [29] and pigs, and turkey in Great Britain [30]. In Iran, the meat industry is mostly based on poultry producing 2.2 million tons yearly being the seventh largest producer of chicken meat in the world [31]. On the other hand, pig is not bred in house-hold or in industry and there is no record of importing the pork products. This condition could eliminate the propagation and circulation of the well adapted monophasic S.1,4, [5],12:i:-variants of pork origin in Iran. To our knowledge there is no publication on prevalence of monophasic variants from other Islamic countries for comparison.
In this study, 6% (9/150) of the isolates (X, XI) were genotypically non-motile. Two were from red meat and seven from chicken. These isolates were mostly from Tehran (8/9) that were collected in 2015. In the study by Bugarel et al., the rate of non-motile was close to 6.6% (17/256), but none of the isolates were genotypically negative for both genes [19].
Some of the factors that affect the severity and prognosis of the disease induced by S.Typhimurium are the presence of DT104 prophage [33], flagellar antigens and IS200 [25]. S.Typhimurium DT104 is a human food-borne and animal pathogen that carries DT104 phage. In our study, 150 samples were tested for DT104, the highest isolation rate was observed in samples from cattles (86.3% (19/22) Table 5), highlighting the importance of these in dessimination of DT104/U302 subtypes with possible phage-associated resistance genes. A review in 1995 from UK has shown that the prevalence of S.Typhimurium DT104 in cattle/sheep, swine and poultry was 70%, 36%, and 52%, respectively [34]. However, in our study, the prevalence of DT104 for cattle/sheep, human and chicken were 86.3%, 68.8%, and 42.2%, respectively. Almost half the S. Typhimurium isolates (45.1% (42/93) in group I or in classic group belonged to this subtype (Table 8).
Flagella play an important role as receptor for DT104 [22] and among the 65 confirmed DT104 positive isolates 64.6% (42/65) carried both flagellar genes; 18.6% (12/65) were positive for fljB and negative for fliC gene and 9.2% (6/65) were fliC positive and fljB negative (Table 8). This is in accordance with Shin et al. that showed the presence of both flagella is important as receptor for DT104 phage that belongs to the group FI and FII of phage families [22].
One of the important markers of S. Typhimurium is IS200 sequence located in fliA-B region, downstream of fliC gene [35]. According to the EFSA guidelines for S. Typhimurium identification, both classical and fljBmonophasic variants carry IS200 in fliB-fliA intergenic region by which S.Typhimurium and its variant are distinguished from other Salmonella serovars namely: S. Lagos, S. Agama, S. Farsta, S. Tsevie, S. Gloucester, and S. Tumodi [10]. Here in human category (Table 4), 3/13 classical and 1/6 monophasicS.1,4, [5],12:i:-of confirmed S. Typhimurium serovar showed deletion of IS200. Moreover, 12.7% (19/150) of all isolated S. Typhimurium in this study (Table 8) lacked IS200 and thus did not support the EFSA suggestion that IS200 should be present in all variants of S.Typhimurium. Bugarel et al has also shown that IS200 is absent in 0.4% (1/256) of the European isolates [19] Beuzone et al. have shown that IS200 could rearrange chromosome up to several kb near its insertion site when it transposes [35]. The proximity (<1.5 kb) of the fliC gene and IS200 insertion site, may also affect the presence of fliC gene when IS200 transposes. In our study, the comparison of the ratio of S.Typhimurium lacking IS200 among variant I (4.3%) and variant VII (25.7%) showed 5.97 times increase in the absence of IS200 in variant VII that lacks fliC gene (Table 8).

Conclusion
Our results showed that S. Typhimurium with IS200deletion is a fairly common serotype among the Iranian Salmonella isolates of human and non-human origins. Therefore, with the procedure recommended by EFSA some of the variants will not be correctly identified and therefore missed. In addition, with 34.3% prevalence of DT104 subtype that could carry resistant genes, among the atypical strains highlight the importance of adaptation of new guidelines for identification of all the variants.