In sub-Saharan Africa, non-typhoidal Salmonella (NTS) are emerging as a prominent cause of invasive disease (bacteremia and focal infections such as meningitis) in infants and young children. Importantly, including data from Mali, three serovars, Salmonella enterica serovar Typhimurium, Salmonella Enteritidis and Salmonella Dublin, account for the majority of non-typhoidal Salmonella isolated from these patients.
We have extended a previously developed series of polymerase chain reactions (PCRs) based on O serogrouping and H typing to identify Salmonella Typhimurium and variants (mostly I 4,,12:i:-), Salmonella Enteritidis and Salmonella Dublin. We also designed primers to detect Salmonella Stanleyville, a serovar found in West Africa. Another PCR was used to differentiate diphasic Salmonella Typhimurium and monophasic Salmonella Typhimurium from other O serogroup B, H:i serovars. We used these PCRs to blind-test 327 Salmonella serogroup B and D isolates that were obtained from the blood cultures of febrile patients in Bamako, Mali.
We have shown that when used in conjunction with our previously described O-serogrouping PCR, our PCRs are 100% sensitive and specific in identifying Salmonella Typhimurium and variants, Salmonella Enteritidis, Salmonella Dublin and Salmonella Stanleyville. When we attempted to differentiate 171 Salmonella Typhimurium (I 4,[ 5],12:i:1,2) strains from 52 monophasic Salmonella Typhimurium (I 4,,12:i:-) strains, we were able to correctly identify 170 of the Salmonella Typhimurium and 51 of the Salmonella I 4,,12:i:- strains.
The genus Salmonella has more than 2500 serological variants (serovars), such as Salmonella enterica serovar Typhi and Salmonella Paratyphi A and B, that cause, respectively, typhoid and paratyphoid fevers (enteric fevers), and a large number of non-typhoidal Salmonella (NTS) serovars that cause gastroenteritis in healthy hosts. In young infants, the elderly and immunocompromised hosts, NTS can cause severe, fatal invasive disease. Multiple studies of pediatric patients in sub-Saharan Africa have documented the important role of NTS, in particular Salmonella Typhimurium and Salmonella Enteritidis (and to a lesser degree Salmonella Dublin), as invasive bacterial pathogens. Salmonella spp. are isolated from blood and identified by standard microbiological techniques and the serovar is ascertained by agglutination with commercial antisera. PCR-based typing techniques are becoming increasingly popular in developing countries, in part because high quality typing sera are difficult to obtain and expensive and H serotyping is technically difficult. We have developed a series of polymerase chain reactions (PCRs) to identify Salmonella Typhimurium and variants, Salmonella Enteritidis and Salmonella Dublin. We successfully identified 327 Salmonella isolates using our multiplex PCR. We also designed primers to detect Salmonella Stanleyville, a serovar found in West Africa. Another PCR generally differentiated diphasic Salmonella Typhimurium and monophasic Salmonella Typhimurium variant strains from other closely related strains. The PCRs described here will enable more laboratories in developing countries to serotype NTS that have been isolated from blood.
Citation: Tennant SM, Diallo S, Levy H, Livio S, Sow SO, Tapia M, et al. (2010) Identification by PCR of Non-typhoidal Salmonella enterica Serovars Associated with Invasive Infections among Febrile Patients in Mali. PLoS Negl Trop Dis 4(3): e621. https://doi.org/10.1371/journal.pntd.0000621
Editor: Sharon J. Peacock, Cambridge University, United Kingdom
Received: October 6, 2009; Accepted: January 20, 2010; Published: March 9, 2010
This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose.
Funding: This work was supported by grants from the Bill & Melinda Gates Foundation, the Optimus Foundation and the National Institute of Allergy and Infectious Diseases (NIAID), USA. The Gates Foundation grants (#1187 and #32470, to MML) supported surveillance to detect invasive bacterial infections in infants and children (in particular, Haemophilus influenzae type b [Hib]) and to see the impact of introduction of Hib conjugate on invasive Hib disease (#32470). The grant from the Optimus Foundation (to SOS) supported systematic blood culture surveillance to detect non-typhoidal Salmonella. One grant from NIAID (R01 AI029471 to MML) supports the use of attenuated Salmonella to the immune system. A second grant from NIAID (U54 AI057168-06 to MML) supports the development of a vaccine to prevent invasive non-typhoidal Salmonella disease. These funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
In industrialized countries, non-typhoidal Salmonella (NTS) constitute a well recognized public health problem that in healthy subjects is overwhelmingly encountered clinically as self-limited gastroenteritis ,. In immunocompromised and debilitated hosts, NTS can become invasive, leading to bacteremia, sepsis and focal infections (e.g., meningitis) ,. Among infants less than three months of age who become infected with NTS in industrialized countries, invasiveness is also occasionally observed, resulting in bacteremia and focal infections .
Interestingly, whereas systematic blood culture-based surveillance of febrile pediatric patients in Asia has clearly highlighted the high incidence of bacteremia associated with Salmonella enterica serovars Typhi and Paratyphi A in children residing in crowded urban settings –, isolation of NTS has not been common. In striking contrast, systematic blood culture-based surveillance and clinical studies of hospitalized and ambulatory pediatric patients <60 months of age with fever or focal infections in sub-Saharan Africa have documented the important role of NTS as invasive bacterial pathogens –. NTS constituted one of the three most common invasive bacterial pathogens in all these studies. Importantly, two serovars, Salmonella Typhimurium (and Typhimurium variants) and Salmonella Enteritidis have been reported to account for 79–95% of all bacteremic non-typhoidal Salmonella infections in sub-Saharan Africa , –,,,,. Salmonella Dublin has been associated with a few percent of cases in some studies , but with a more substantial proportion in Mali , where a fourth serovar, Salmonella Stanleyville, also accounted for a notable proportion of all isolates , bringing the cumulative total to >95% of all strains.
We previously developed a multiplex polymerase chain reaction (PCR)-based approach to identify the three main pathogens responsible for typhoid (Salmonella Typhi) and paratyphoid (Salmonella Paratyphi A and Salmonella Paratyphi B) fevers . Three sequential PCRs identify strains of Salmonella serogroups A, B or D (and Vi positive or negative); strains that express Phase 1 flagellar (H) antigen types H:a, H:b or H:d; and strains incapable of fermenting d-tartrate (d-T). By means of this PCR technology, Salmonella Typhi (O serogroup D, Vi+; H:d), Salmonella Paratyphi A (O serogroup A; H:a) and Salmonella Paratyphi B (O serogroup B; H:b; d-T non-fermenter) strains were identified with 100% sensitivity and 100% specificity.
Classical Salmonella serotyping methods identified the serovars of 336 NTS isolates from blood cultures of febrile children <16 years of age in Bamako, Mali, obtained in the course of systematic surveillance of children admitted to hospital or seen in the Emergency Department with fever or invasive infection syndromes –. Salmonella Typhimurium and “variants” (mainly I 4,,12:i:-), Salmonella Dublin, Salmonella Enteritidis and Salmonella Stanleyville were the most commonly isolated NTS . Herein, we describe PCRs that when used in conjunction with the O serogrouping PCR described by Levy et al.  can identify Salmonella Typhimurium and variants (O serogroup B; H:i), Salmonella Enteritidis (O serogroup D, Vi-; H:g,m), Salmonella Dublin (O serogroup D, Vi+ or Vi-; H:g,p) and Salmonella Stanleyville (O serogroup B; H:z4,z23) with 100% sensitivity and 100% specificity. We anticipate that this methodology will be useful in reference laboratories and major clinical microbiology laboratories to identify Salmonella isolated from blood and other sterile sites in developing countries where robust PCR-based typing techniques are becoming increasingly popular and because high quality H typing sera are difficult to obtain, expensive and technically demanding to use.
The surveillance protocol and consent form were reviewed by the Ethics Committee of the Faculté de Médecine, Pharmacie et Odonto-Stomatologie, Université de Bamako, and by the Institutional Review Board of the University of Maryland, Baltimore. For any patient eligible for laboratory surveillance to detect invasive bacterial disease, informed consent was obtained prior to their enrollment; ∼95% of eligible subjects agreed to participate. Since the literacy rate in Bamako is <30%, as is customary practice for CVD-Mali clinical studies –, the consent form was translated into Bambara and several other local languages and the translations recorded on audiotape . CVD-Mali personnel explain the study, including the objectives and risks and benefits associated with participation. The audiotaped version of the consent form is then played and any questions posed are answered. Once the parent or patient has had all questions answered and agrees to participate, this is documented on a printed consent form written in French. If the participant is illiterate, a witness who is present throughout the consent procedure completes the necessary portions and signs the consent form; the parent/participant marks the consent form (either fingerprint or other notation). If the person is literate, then he/she may read and sign the consent form. This standard method of obtaining consent practiced by CVD-Mali was approved by ethics commitees in Mali and at the University of Maryland.
Systematic surveillance for invasive bacterial infections
Since July 2002, clinical staff of the Centre pour le Développement des Vaccins du Mali (CVD-Mali) and l'Hôpital Gabriel Touré (HGT) have been conducting systematic surveillance to detect invasive bacterial disease among hospitalized children <16 years of age –. Age-eligible children presenting to the emergency department with fever (≥39°C) or focal clinical findings suggestive of invasive bacterial infection (meningitis, septic arthritis, etc.) and requiring hospitalization are referred to CVD-Mali staff by the evaluating clinicians. A CVD-Mali physician obtains informed consent, records clinical and epidemiologic data, and obtains blood (and other relevant fluids) for culture in the HGT Clinical Bacteriology Laboratory. The child's clinician is promptly notified when a culture yields a bacterial pathogen.
Salmonella Typhimurium strain 81.23500, Salmonella Enteritidis strain CVD SE and Salmonella Dublin strain 06-0707 were used to develop the multiplex PCR. Twenty-four control strains which came from the Salmonella Reference Laboratory of the Centers for Disease Control and Prevention (CDC), Atlanta, GA or the Center for Vaccine Development, Baltimore, MD have previously been described . These strains were Salmonella serovars of various O serogroups (B, C1, C2, D, E1, O28 and O38) and H types (b, c, d, h, i, g, k, l, m, p, s, t, v, y, z10 and z29). Nine O serogroup B, Phase 1 flagella antigen H:i reference strains from the CDC were used to develop a PCR that discriminates between Salmonella Typhimurium and I 4,,12:i:- (Table 1). The NTS test strains consist of 327 Salmonella serogroup B and D isolates that were originally obtained from the blood cultures of febrile patients at l'Hôpital Gabriel Touré in Bamako, Mali. These strains were identified by conventional microbiological and classical serotyping methods by the CVD and CDC, as previously described ; 69 isolates were O serogroup D, including 37 Salmonella Dublin and 32 Salmonella Enteritidis, and 258 isolates were O serogroup B.
Detection of Salmonella Typhimurium and variants, Salmonella Enteritidis and Salmonella Dublin (and later Salmonella Stanleyville).
This primer mix contained the following primers: H-for, a primer sequence common to fliC of both Salmonella Typhimurium (H:i) and Salmonella Dublin (H:g,p); Hi, unique to fliC of H:i organisms; Hgp, unique to fliC of H:g,p organisms; sdfF and sdfR, which amplify Sdf I, a fragment of DNA unique to Salmonella Enteritidis; and 16SF and DG74, universal bacterial 16S rRNA gene primers that were included to ensure that DNA was added (Table 2).
Primer Hgp was designed by performing Clustal W alignments between fliC nucleotide sequences deposited in GenBank (three Salmonella Enteritidis sequences, DQ095560, AY649709 and AY649742; one Salmonella Dublin sequence, AY649712). The fliC alleles of Salmonella Enteritidis (fliC-g,m) and Salmonella Dublin (fliC-g,p) are very similar, as they are both G complex alleles that are almost identical to one another . The nucleotide sequence of fliC-g,p, to which primer Hgp binds, differs from fliC-g,m by two nucleotides. The Hgp primer was therefore designed such that one of these mismatches was at the 3′ end of Hgp (fliC-g,p has a ‘T’ whereas fliC-g,m has a ‘C’).
In a later step, primers to detect fliC of Salmonella Stanleyville (H:z4,z23) were added to the primer mix. Primers Hz4,z23F and Hz4,z23R were designed to amplify a 427-bp fragment of fliC-z4,z23 (based on GenBank accession no. AY649736).
Differentiation of diphasic Salmonella Typhimurium and monophasic Salmonella Typhimurium from other H:i serovars.
We combined published primers in a PCR to discriminate between Salmonella Typhimurium (I 4,,12:i:1,2) and the monophasic serovar I 4,,12:i:- and other O serogroup B, H:i serovars. To accomplish this, we used a primer mix containing primers FFLIB and RFLIA, which amplify the fliB-fliA intergenic region, and primers Sense-59 and Antisense-83, which amplify the Phase 2 (fljB) flagellar gene.
DNA extraction and PCR
PCR was performed in 1× PCR buffer, 3.5 mM MgCl2, 0.2 mM of dNTPs and 0.2 U of Invitrogen Taq DNA polymerase (final volume of 25 µl) in an Eppendorf Mastercycler®. The primer mixes contained primers at a concentration of 5 µM each (final concentration of 0.2 µM) except for FFLIB and RFLIA that were used at a concentration of 10 µM each and the positive control primers (16SF and DG74) that were used at a concentration of 2.5 µM each. For each PCR reaction, 1.0 µl of primer mix was used. Crude DNA was prepared by suspending 3 colonies in 100 µl water and boiling for 10 min followed by centrifugation at 16,000×g for 30 sec and purified DNA was prepared using a GNOME DNA kit (QBIOgene, Irvine, CA) according to the manufacturer's instructions, and 5 µl of DNA was used in each PCR. The cycling parameters of the multiplex PCR that detects H:i, H:g,p and Sdf I involved denaturation at 95°C for 2 min, followed by 25 cycles comprised of heating to 95°C for 30 sec, 64°C for 30 sec and 72°C for 15 sec, and a final step of 72°C for 5 min. The cycling parameters of the PCR that discriminates between Salmonella Typhimurium and I 4,,12:i:- involved denaturation at 95°C for 2 min, followed by 25 cycles of 95°C for 30 sec, 64°C for 30 sec and 72°C for 1.5 min, and a final step of 72°C for 5 min. PCR products were separated on 2% (w/v) agarose gels, stained with ethidium bromide and visualized using a UV transilluminator.
A multiplex PCR to detect Salmonella Typhimurium and monophasic variants, Salmonella Enteritidis and Salmonella Dublin
Figure 1 shows that the primers within the multiplex PCR were able to clearly identify the appropriate NTS alleles. A 779-bp product was amplified from Salmonella Dublin (fliC-gp), a 551-bp product was amplified from Salmonella Typhimurium (fliC-i) and a 333-bp product was amplified from Salmonella Enteritidis (Sdf I). The internal positive control primers (universal 16S rRNA gene primers) amplified a 167-bp product from each strain.
Initial validation of the multiplex PCR assay
To preliminarily assess the specificity of the multiplex PCR assay, we tested 24 control Salmonella strains consisting of a range of serovars (previously described in ) in a blinded fashion (Figure 2). The multiplex PCR correctly identified Salmonella Typhimurium and Salmonella Cotham as H:i, Salmonella Dublin as H:g,p and Salmonella Enteritidis as containing Sdf I (Figure 2). Faint products of the size of Sdf I were observed for Salmonella Meleagridis and Salmonella Livingstone. However, Salmonella Meleagridis is O serogroup E1 and Salmonella Livingstone is O serogroup C1, so when also tested by our previously described O serogrouping PCR , these serovars would not be mistaken as Salmonella Enteritidis. The same is true for Salmonella Cotham, which although it possesses fliC-i, is not O serogroup B and would not be mistaken as Salmonella Typhimurium. Therefore, the new multiplex PCR was sensitive in terms of its ability to identify serovar Cotham as H:i and was specific, when combined with the O-serogrouping PCR, in showing that the strain was not serovar Typhimurium. We also blind-tested a sample of Salmonella Typhi and Salmonella Paratyphi A and B strains to ensure that the PCR would not detect these strains. The multiplex PCR correctly identified fliC-i of six Salmonella Typhimurium, Sdf I of four Salmonella Enteritidis, and fliC-g,p of five Salmonella Dublin strains but only the 16S rRNA gene was amplified from five strains each of serovars Typhi, Paratyphi A and Paratyphi B (data not shown).
Reference strains consisted of one representative each of Salmonella Typhimurium, Salmonella Enteritidis and Salmonella Dublin and a range of negative control Salmonella strains of various serovars. PCR controls consisted of Salmonella Typhimurium 81.23500, Salmonella Enteritidis CVD SE, Salmonella Dublin 06-0707.
Analysis of 327 Salmonella seroroup B and D clinical isolates from Mali
We blind-tested 69 non-Typhi serogroup D Salmonella and 258 serogroup B strains that were originally obtained from the blood cultures of febrile patients at l'Hôpital Gabriel Touré in Bamako, Mali  with the multiplex PCR designed to identify Salmonella Typhimurium (I 4,,12:i:1,2) and variants (monophasic I 4,,12:i:- and non-motile (NM) I 4,,12:NM), Salmonella Enteritidis and Salmonella Dublin. This PCR was performed in parallel to serotyping. We correctly identified all the serogroup D isolates (37 Salmonella Dublin and 32 Salmonella Enteritidis) and all 232 Salmonella Typhimurium and variant strains (Table 3). If the Salmonella Typhimurium-like strains (i.e., I 4,,12:i:- and I 4,,12:NM) are included in the target group then the PCR is 100% sensitive and 100% specific in identifying Salmonella Typhimurium, Salmonella Enteritidis, Salmonella Dublin and Salmonella Typhimurium-like organisms. The remaining 26 serogroup B isolates were negative for the tested targets.
Discrimination between Salmonella Typhimurium and I 4,,12:i:-
During the course of this study, we decided to determine the prevalence of I 4,,12:i:- in Mali. Levy et al.  identified 220 Salmonella Typhimurium, four I 4,,12:i:- and eight I 4,,12:NM strains. However, in this previous study, Phase 2 flagella typing was not performed on all of the strains. We re-examined the 220 Salmonella O serogroup B, H:i isolates that had been previously been presumptively identified as Salmonella Typhimurium and used classical methods (i.e., sera against the Phase 2 H1,2 flagella) to determine that 48 isolates were in fact I 4,,12:i:- (bringing the total number of isolates of this serovar to 52) and one isolate was I 4,,12:NM (bringing the total number of isolates of this serovar to nine). The remaining 171 strains were confirmed as Salmonella Typhimurium.
We have combined previously described primers in a PCR to discriminate between Salmonella Typhimurium and I 4,,12:i:-. Primers FFLIB and RFLIA amplify the fliB-fliA intergenic region of the flagellin gene cluster . Salmonella Typhimurium strains possess an IS200 fragment in this region . Burnens et al.  showed that 21 of 23 isolates of Salmonella Typhimurium and none of 85 isolates of 37 other Salmonella serovars contained IS200 in this region. Primers FFLIB and RFLIA have been reported to amplify a 1-kb product from Salmonella Typhimurium and I 4,,12:i:- strains and a 250-bp product from all other serovars . However, when validating these primers, we found that a 1-kb fragment was amplified from Salmonella Farsta (not tested by Echeita et al. ) suggesting that this serovar also possesses IS200 in the fliB-fliA intergenic region (Figure 3).
The PCR to discriminate between Salmonella Typhimurium and I 4,,12:i:- strains was validated using nine reference Salmonella O group B, H:i isolates of various serovars.
Primers Sense-59 and Antisense-83 amplify the fljB allele . Primer Sense-59 binds at position 258 and primer Antisense-83 binds at position +100 of the 5′-3′ consensus fljB1,2 sequence. These primers amplify a 1389-bp product from strains that possess a Phase 2 flagellar antigen and no product from strains that lack a Phase 2 flagellar antigen such as I 4,,12:i:-. As shown in Figure 3, the PCR was able to discriminate between Salmonella Typhimurium and I 4,,12:i:- strains and other serogroup B, H:i serovars except Salmonella Farsta.
We tested all the Salmonella Typhimurium, I 4,,12:i:- and I 4,,12:NM strains identified in Mali and found that 170 of 171 Salmonella Typhimurium strains were correctly identified (i.e., possessed a 1-kb fliB-fliA intergenic region product and fljB1,2), and 51 of 52 I 4,,12:i:- strains were correctly identified (i.e., possessed a 1-kb fliB-fliA intergenic region product and lacked fljB1,2) (Table 4). The nine I 4,,12:NM strains produced mixed results in that all nine strains produced a 1-kb fliB-fliA intergenic region product but three strains possessed fljB1,2.
Detection of Salmonella Stanleyville
Since Salmonella Stanleyville was found to be fairly common among the Mali NTS isolates, we added primers to detect fliC-z4,z23 of Salmonella Stanleyville to the multiplex PCR containing primers H-for, Hi, sdfF, sdfR, 16SF and DG74. The primers were first tested on Salmonella Stanleyville by themselves and produced a 427-bp amplicon. The fliC-z4,z23 primers were then added to the multiplex primer mix and PCR was performed (using the previously optimized conditions) on all 26 Salmonella Stanleyville strains, and a sample of 10 Salmonella Typhimurium, 10 Salmonella Dublin and 11 Salmonella Enteritidis strains. Correct amplicons were observed for all the strains tested. Figure 4 shows amplicons from a sample of three Salmonella Stanleyville strains and the control Salmonella Typhimurium, Salmonella Enteritidis and Salmonella Dublin strains.
Amplification of fliC from 3 Salmonella Stanleyville strains from Mali and Salmonella Typhimurium 81.23500, Salmonella Enteritidis CVD SE and Salmonella Dublin 06-0707 following the addition of Hz4,z23 primers to the fliC-i/fliC-gp/Sdf I multiplex PCR mix.
We have combined published primers and new primers in a multiplex PCR that, following the application of a previously described O serogrouping multiplex PCR , can identify Salmonella Typhimurium (and variants), Salmonella Enteritidis, Salmonella Dublin and Salmonella Stanleyville. Detection of Salmonella Typhimurium, Salmonella Dublin and Salmonella Stanleyville is based on amplification of the respective fliC alleles. We were unable to design primers to detect fliC-g,m of Salmonella Enteritidis due to the high nucleotide identity between fliC-g,m and fliC-g,p (of Salmonella Dublin). We therefore used primers to detect “Salmonella difference fragment I” (Sdf I), a segment of Salmonella Enteritidis DNA that was reported to be absent from 73 non-Enteritidis Salmonella enterica isolates comprising 34 different serovars as determined by PCR . We confirmed the utility of Sdf I, with the exception of serovars Meleagridis and Livingstone. We found that Salmonella Livingstone yielded a weak PCR product using the same Sdf I primers that were previously reported . The disparity could be due to a difference in the amplification method (different polymerases and cycling conditions were used).
From the epidemiologic and public health perspective, being able to detect strains that are genetically similar to Salmonella Typhimurium yet that constitute distinct serovars (i.e., I 4,,12:i:- and I 4,,12:NM) is important (e.g., for outbreak investigations). In the USA and Europe such strains are increasingly being reported –. In Spain, I 4,,12:i:- was the fourth most commonly isolated Salmonella serovar from humans from 1998–1999  and several studies suggest that this monophasic serovar is a variant of Salmonella Typhimurium , –. The PCR that we have described can generally discriminate the diphasic Salmonella Typhimurium serovar (I 4,,12:i:1,2) from monophasic (I 4,,12:i:-) variants. Only one Salmonella Typhimurium was misidentified as I 4,,12:i:- and vice versa. It is possible that our PCR will not be able to detect some serologically monophasic I 4,,12:i- strains as lack of Phase 2 flagellar antigen expression can be due to a variety of mechanisms ranging from point mutations to partial or complete deletions in fljB1,2 and adjacent genes. Additionally, if there is a deletion in the first 250 bp of fljB1,2, the primers we have chosen will not identify the strain as I 4,,12:i-. Furthermore, our PCR scheme cannot differentiate between Salmonella Typhimurium and Salmonella Farsta. However, in practical terms, this is unlikely to pose a problem as Salmonella Farsta is extremely rare.
One small set of strains where our PCR gives differing results from traditional serological methods are Salmonella Typhimurium-like non-motile variants (I 4,,12:NM). Notably, all nine Malian strains identified by serotyping methods as I 4,,12:NM were found to possess the fliC-i allele and three of the strains also possessed the fljB1,2 gene. Two quite distinct explanations can account for these observations. One is that in some strains lack of motility is not due to loss of flagellar genes but rather to other factors (e.g., regulation) that keep expression turned off. Alternatively, it may be that our genetic identification of these strains is correct and that the failure to detect flagella phenotypically is merely a consequence of not knowing how to grow the bacteria under conditions optimal for expression of those flagella. We assume that the I 4,,12:NM strains from Mali are Salmonella Typhimurium variants as they possess fliC-i and IS200 in the fliB-fliA intergenic region. It is also possible, albeit unlikely, that they could be the very rarely isolated Salmonella Farsta.
Soyer et al.  have reported that there are at least two common clones of I 4,,12:i:- with different genomic deletions (an ‘American’ deletion genotype and a ‘Spanish’ deletion genotype). Both I 4,,12:i:- clones completely lack fljB and fljA. Preliminary analysis of the deletion using a variety of primers that amplify different sections of the fljB1,2 gene indicates that the I 4,,12:i:- strains from Mali appear to possess the 3′ end of fljB and the entire fljA ORF. At least 250 bp of fljB (including the Sense-59 binding site) has been deleted at the 5′ end (data not shown). This suggests that these strains are genetically different from both the Spanish and American I 4,,12:i:- isolates. We are sequencing the deletion in several Malian I 4,,12:i- strains to determine the exact deletion. It will be interesting to see whether I 4,,12:i:- strains from other African countries are genetically similar to the Malian strains.
Several other DNA-based Salmonella typing methods have been described –. However, some of these do not identify the breadth of enteric fever and NTS serovars of our multistep, multiplex PCR or fail to include an internal positive control. An O serogroup-specific Bio-Plex assay to detect serogroups B, C1, C2, D, E and O13 and serovar Paratyphi A  and a DNA sequence-based approach to serotyping have also been described . However, these methods require greater financial and technical resources over those required for our method. Our PCRs are novel because they use as few primers as possible to identify the most common non-typhoidal Salmonella serovars isolated from blood and other invasive sites in sub-Saharan Africa, including Salmonella Typhimurium (and several variants), Salmonella Enteritidis, Salmonella Dublin and Salmonella Stanleyville. Since the late 1980s, the majority (85 to 95%) of NTS associated with invasive disease in sub-Saharan Africa belong to these serovars , –,,,,. Therefore, we do not believe that there is a need for multiplex PCRs that detect more serovars unless the epidemiologic picture changes. We have tried to keep the PCRs as simple as possible so that they can be performed easily and the results interpreted correctly in laboratories in Africa that may be new to PCR. If a large outbreak or otherwise frequent isolation occurred of a serovar not presently recognized or contained within our multiplex, this serovar would not be identifiable using our PCR and would have to be identified in a reference laboratory using antisera or by molecular serotyping.
We are currently evaluating various PCR reagents that are stable at room-temperature and can be readily obtained by laboratories in Africa. Depending on the prevalence of certain serovars in a given country, either typhoidal or non-typhoidal Salmonella (or both) can be identified using our primer sets (Table 5 and Figure 5). For example, one may wish to test all Salmonella isolates in the O serogrouping PCR, then screen serogroup A, B and D Vi+ strains using the first H typing multiplex PCR to identify Salmonella Typhi, Salmonella Paratyphi A and Salmonella Paratyphi B. The d-tartrate fermentation PCR can be performed to differentiate Salmonella Paratyphi B sensu stricto strains from Salmonella Paratyphi B Java. Any serogroup B isolates not identified by the 1st H typing PCR can be tested along with non-Typhi O serogroup D strains in the second H typing/Sdf I multiplex PCR to identify serovars Typhimurium (and related strains), Dublin (which can be Vi+ or Vi- ), Enteritidis and Stanleyville. The O serogroup B H:i strains can be tested using the Typhimurium/I 4,,12:i:- PCR to identify Salmonella Typhimurium and I 4,,12:i:-. It should be stressed that the O serogrouping PCR described by Levy et al.  needs to be performed in conjunction with the PCRs described here to ensure that Salmonella Enteritidis and Salmonella Typhimurium are identified correctly and not mistaken as Salmonella Meleagridis and Salmonella Livingstone; and Salmonella Cotham, respectively.
The surveillance experience in Mali is the first to show that Salmonella Dublin and Salmonella Stanleyville can constitute important serovars associated with invasive non-typhoidal Salmonella disease, along with Salmonella Typhimurium (and variants) and Salmonella Enteritidis. Previously, Salmonella Dublin and Salmonella Stanleyville were recovered only occasionally from blood cultures of patients in Africa ,,,. We thought it useful to be able to detect these serovars by PCR in future surveillance studies in Africa.
In conclusion, we have described a series of PCRs based on O serogrouping and H typing that can identify the causative agents of enteric fever (Salmonella Typhi and Salmonella Paratyphi A and Salmonella Paratyphi B), the three most commonly isolated serovars that cause invasive disease in young children in sub-Sahara African (Salmonella Typhimurium [and Typhimurium-like], Salmonella Enteritidis and Salmonella Dublin) and Salmonella Stanleyville, an invasive pathogen that may be of regional importance in West Africa.
Conceived and designed the experiments: SMT HL MML. Performed the experiments: SMT SD HL SL SOS MT MM BT KLK JEG. Analyzed the data: SMT SL SOS MT PIF BT JPN MML. Contributed reagents/materials/analysis tools: SD HL SL SOS PIF MM KLK JPN JEG. Wrote the paper: SMT MML.
- 1. Voetsch AC, Van Gilder TJ, Angulo FJ, Farley MM, Shallow S, et al. (2004) FoodNet estimate of the burden of illness caused by nontyphoidal Salmonella infections in the United States. Clin Infect Dis 38: Suppl 3S127–S134.
- 2. Adak GK, Long SM, O'Brien SJ (2002) Trends in indigenous foodborne disease and deaths, England and Wales: 1992 to 2000. Gut 51: 832–841.
- 3. Kennedy M, Villar R, Vugia DJ, Rabatsky-Ehr T, Farley MM, et al. (2004) Hospitalizations and deaths due to Salmonella infections, FoodNet, 1996-1999. Clin Infect Dis 38: Suppl 3S142–S148.
- 4. Vugia DJ, Samuel M, Farley MM, Marcus R, Shiferaw B, et al. (2004) Invasive Salmonella infections in the United States, FoodNet, 1996-1999: incidence, serotype distribution, and outcome. Clin Infect Dis 38: Suppl 3S149–S156.
- 5. Brooks WA, Hossain A, Goswami D, Nahar K, Alam K, et al. (2005) Bacteremic typhoid fever in children in an urban slum, Bangladesh. Emerg Infect Dis 11: 326–329.
- 6. Phetsouvanh R, Phongmany S, Soukaloun D, Rasachak B, Soukhaseum V, et al. (2006) Causes of community-acquired bacteremia and patterns of antimicrobial resistance in Vientiane, Laos. Am J Trop Med Hyg 75: 978–985.
- 7. Sinha A, Sazawal S, Kumar R, Sood S, Reddaiah VP, et al. (1999) Typhoid fever in children aged less than 5 years. Lancet 354: 734–737.
- 8. Berkley JA, Lowe BS, Mwangi I, Williams T, Bauni E, et al. (2005) Bacteremia among children admitted to a rural hospital in Kenya. N Engl J Med 352: 39–47.
- 9. Brent AJ, Oundo JO, Mwangi I, Ochola L, Lowe B, et al. (2006) Salmonella bacteremia in Kenyan children. Pediatr Infect Dis J 25: 230–236.
- 10. Graham SM, Walsh AL, Molyneux EM, Phiri AJ, Molyneux ME (2000) Clinical presentation of non-typhoidal Salmonella bacteraemia in Malawian children. Trans R Soc Trop Med Hyg 94: 310–314.
- 11. Kariuki S, Revathi G, Kariuki N, Kiiru J, Mwituria J, et al. (2006) Characterisation of community acquired non-typhoidal Salmonella from bacteraemia and diarrhoeal infections in children admitted to hospital in Nairobi, Kenya. BMC Microbiol 6: 101.
- 12. Ikumapayi UN, Antonio M, Sonne-Hansen J, Biney E, Enwere G, et al. (2007) Molecular epidemiology of community-acquired invasive non-typhoidal Salmonella among children aged 2 - 29 months in rural Gambia and discovery of a new serovar, Salmonella enterica Dingiri. J Med Microbiol 56: 1479–1484.
- 13. Lepage P, Bogaerts J, Van Goethem C, Ntahorutaba M, Nsengumuremyi F, et al. (1987) Community-acquired bacteraemia in African children. Lancet 1: 1458–1461.
- 14. O'Dempsey TJ, McArdle TF, Lloyd-Evans N, Baldeh I, Laurence BE, et al. (1994) Importance of enteric bacteria as a cause of pneumonia, meningitis and septicemia among children in a rural community in The Gambia, West Africa. Pediatr Infect Dis J 13: 122–128.
- 15. Walsh AL, Phiri AJ, Graham SM, Molyneux EM, Molyneux ME (2000) Bacteremia in febrile Malawian children: clinical and microbiologic features. Pediatr Infect Dis J 19: 312–318.
- 16. Sigauque B, Roca A, Mandomando I, Morais L, Quinto L, et al. (2009) Community-acquired bacteremia among children admitted to a rural hospital in Mozambique. Pediatr Infect Dis J 28: 108–113.
- 17. Mandomando I, Macete E, Sigauque B, Morais L, Quinto L, et al. (2009) Invasive non-typhoidal Salmonella in Mozambican children. Trop Med Int Health 14: 1467–1474.
- 18. Levy H, Diallo S, Tennant SM, Livio S, Sow SO, et al. (2008) A PCR Method to Identify Salmonella enterica serovars Typhi, Paratyphi A and Paratyphi B Among Salmonella Isolates from the Blood of Patients with Clinical Enteric Fever. J Clin Microbiol 46: 1861–1866.
- 19. Kariuki S, Revathi G, Kariuki N, Kiiru J, Mwituria J, et al. (2006) Invasive multidrug-resistant non-typhoidal Salmonella infections in Africa: zoonotic or anthroponotic transmission? J Med Microbiol 55: 585–591.
- 20. Sow SO, Diallo S, Campbell JD, Tapia MD, Keita T, et al. (2005) Burden of invasive disease caused by Haemophilus influenzae type b in Bamako, Mali: impetus for routine infant immunization with conjugate vaccine. Pediatr Infect Dis J 24: 533–537.
- 21. Campbell JD, Kotloff KL, Sow SO, Tapia M, Keita MM, et al. (2004) Invasive pneumococcal infections among hospitalized children in Bamako, Mali. Pediatr Infect Dis J 23: 642–649.
- 22. Sow SO, Tapia MD, Diallo S, Keita MM, Sylla M, et al. (2009) Haemophilus influenzae Type B conjugate vaccine introduction in Mali: impact on disease burden and serologic correlate of protection. Am J Trop Med Hyg 80: 1033–1038.
- 23. McQuiston JR, Parrenas R, Ortiz-Rivera M, Gheesling L, Brenner F, et al. (2004) Sequencing and comparative analysis of flagellin genes fliC, fliB, and flpA from Salmonella. J Clin Microbiol 42: 1923–1932.
- 24. Echeita MA, Herrera S, Usera MA (2001) Atypical, fljB-negative Salmonella enterica subsp. enterica strain of serovar 4,5,12:i:- appears to be a monophasic variant of serovar Typhimurium. J Clin Microbiol 39: 2981–2983.
- 25. Burnens AP, Stanley J, Sack R, Hunziker P, Brodard I, et al. (1997) The flagellin N-methylase gene fliB and an adjacent serovar-specific IS200 element in Salmonella typhimurium. Microbiology 143 (Pt5): 1539–1547.
- 26. Vanegas RA, Joys TM (1995) Molecular analyses of the phase-2 antigen complex 1,2. of Salmonella spp. J Bacteriol 177: 3863–3864.
- 27. Agron PG, Walker RL, Kinde H, Sawyer SJ, Hayes DC, et al. (2001) Identification by subtractive hybridization of sequences specific for Salmonella enterica serovar Enteritidis. Appl Environ Microbiol 67: 4984–4991.
- 28. Centers for Disease Control and Prevention (2007) Salmonella Surveillance: Annual Summary 2005.Atlanta Department of Health and Human Services.
- 29. Echeita MA, Aladuena A, Cruchaga S, Usera MA (1999) Emergence and spread of an atypical Salmonella enterica subsp. enterica serotype 4,5,12:i:- strain in Spain. J Clin Microbiol 37: 3425.
- 30. Mossong J, Marques P, Ragimbeau C, Huberty-Krau P, Losch S, et al. (2007) Outbreaks of monophasic Salmonella enterica serovar 4,,12:i:- in Luxembourg, 2006. Euro Surveill 12: E11–E12.
- 31. de la Torre E, Zapata D, Tello M, Mejia W, Frias N, et al. (2003) Several Salmonella enterica subsp. enterica serotype 4,5,12:i:- phage types isolated from swine samples originate from serotype typhimurium DT U302. J Clin Microbiol 41: 2395–2400.
- 32. Garaizar J, Porwollik S, Echeita A, Rementeria A, Herrera S, et al. (2002) DNA microarray-based typing of an atypical monophasic Salmonella enterica serovar. J Clin Microbiol 40: 2074–2078.
- 33. Guerra B, Laconcha I, Soto SM, Gonzalez-Hevia MA, Mendoza MC (2000) Molecular characterisation of emergent multiresistant Salmonella enterica serotype [4,5,12:i:-] organisms causing human salmonellosis. FEMS Microbiol Lett 190: 341–347.
- 34. Soyer Y, Moreno SA, Davis MA, Maurer J, McDonough PL, et al. (2009) Salmonella enterica serotype 4,5,12:i:-, an emerging Salmonella serotype that represents multiple distinct clones. J Clin Microbiol 47: 3546–3556.
- 35. Lim YH, Hirose K, Izumiya H, Arakawa E, Takahashi H, et al. (2003) Multiplex polymerase chain reaction assay for selective detection of Salmonella enterica serovar Typhimurium. Jpn J Infect Dis 56: 151–155.
- 36. Luk JM, Kongmuang U, Reeves PR, Lindberg AA (1993) Selective amplification of abequose and paratose synthase genes (rfb) by polymerase chain reaction for identification of Salmonella major serogroups (A, B, C2, and D). J Clin Microbiol 31: 2118–2123.
- 37. Herrera-Leon S, McQuiston JR, Usera MA, Fields PI, Garaizar J, et al. (2004) Multiplex PCR for distinguishing the most common phase-1 flagellar antigens of Salmonella spp. J Clin Microbiol 42: 2581–2586.
- 38. Kim S, Frye JG, Hu J, Fedorka-Cray PJ, Gautom R, et al. (2006) Multiplex PCR-based method for identification of common clinical serotypes of Salmonella enterica subsp. enterica. J Clin Microbiol 44: 3608–3615.
- 39. Hirose K, Itoh K, Nakajima H, Kurazono T, Yamaguchi M, et al. (2002) 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 40: 633–636.
- 40. Herrera-Leon S, Ramiro R, Arroyo M, Diez R, Usera MA, et al. (2007) Blind comparison of traditional serotyping with three multiplex PCRs for the identification of Salmonella serotypes. Res Microbiol 158: 122–127.
- 41. Hong Y, Liu T, Lee MD, Hofacre CL, Maier M, et al. (2008) Rapid screening of Salmonella enterica serovars Enteritidis, Hadar, Heidelberg and Typhimurium using a serologically-correlative allelotyping PCR targeting the O and H antigen alleles. BMC Microbiol. 8:178. 178.
- 42. Fitzgerald C, Collins M, van DS, Mikoleit M, Brown T, et al. (2007) Multiplex, bead-based suspension array for molecular determination of common Salmonella serogroups. J Clin Microbiol 45: 3323–3334.
- 43. Mortimer CK, Peters TM, Gharbia SE, Logan JM, Arnold C (2004) Towards the development of a DNA-sequence based approach to serotyping of Salmonella enterica. BMC Microbiol. 4:31. 31.
- 44. Selander RK, Smith NH, Li J, Beltran P, Ferris KE, et al. (1992) Molecular evolutionary genetics of the cattle-adapted serovar Salmonella dublin. J Bacteriol 174: 3587–3592.
- 45. Bernardo FM, Brandao CF (1996) [Preliminary epidemiological survey on prevalences of Salmonella spp. at Bissau abattoir (Guinea-Bissau)]. Rev Elev Med Vet Pays Trop 49: 102–106.
- 46. Lafaix C, Castets M, Denis F, Diop M, I (1979) [Salmonellosis in Dakar: bacteriological, clinical, epidemiological and therapeutic aspects. Ten years records (author's transl)]. Med Trop (Mars) 39: 369–379.
- 47. Greisen K, Loeffelholz M, Purohit A, Leong D (1994) PCR primers and probes for the 16S rRNA gene of most species of pathogenic bacteria, including bacteria found in cerebrospinal fluid. J Clin Microbiol 32: 335–351.