Staphylococcus aureus is a major human pathogen, a relevant pathogen in veterinary medicine, and a major cause of food poisoning. Epidemiological investigation tools are needed to establish surveillance of S. aureus strains in humans, animals and food. In this study, we investigated 145 S. aureus isolates recovered from various animal species, disease conditions, food products and food poisoning events. Multiple Locus Variable Number of Tandem Repeat (VNTR) analysis (MLVA), known to be highly efficient for the genotyping of human S. aureus isolates, was used and shown to be equally well suited for the typing of animal S. aureus isolates. MLVA was improved by using sixteen VNTR loci amplified in two multiplex PCRs and analyzed by capillary electrophoresis ensuring a high throughput and high discriminatory power. The isolates were assigned to twelve known clonal complexes (CCs) and –a few singletons. Half of the test collection belonged to four CCs (CC9, CC97, CC133, CC398) previously described as mostly associated with animals. The remaining eight CCs (CC1, CC5, CC8, CC15, CC25, CC30, CC45, CC51), representing 46% of the animal isolates, are common in humans. Interestingly, isolates responsible for food poisoning show a CC distribution signature typical of human isolates and strikingly different from animal isolates, suggesting a predominantly human origin.
Citation: Sobral D, Schwarz S, Bergonier D, Brisabois A, Feßler AT, Gilbert FB, et al. (2012) High Throughput Multiple Locus Variable Number of Tandem Repeat Analysis (MLVA) of Staphylococcus aureus from Human, Animal and Food Sources. PLoS ONE 7(5): e33967. doi:10.1371/journal.pone.0033967
Editor: Philip Supply, Institut Pasteur de Lille, France
Received: October 20, 2011; Accepted: February 20, 2012; Published: May 2, 2012
Copyright: © 2012 Sobral et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Work by GV is part of the European Biodefense Laboratory Network (EBLN) project “Database of B-agents” supported by the European Defense Agency. The contribution of SS, ATF and KK is financially supported by the German Federal Ministry of Education and Research (BMBF) through the German Aerospace Center (DLR), grant number 01KI1014D (MedVet-Staph). This study also benefited from the support of the association Vaincre La Mucoviscidose (Grant N° RC0630). The development of tools for the surveillance of bacterial pathogens is supported by the French Direction Générale de l’Armement. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: DS, FLH, and BL are employees of Ceeram and hold stocks. Patent licensing arrangements (Patent number: 1054238; Patent name: Procédé de génotypage de Staphylococcus aureus) exist with DS, FLH, BL, and CP. This does not alter the authors’ adherence to all the PLoS ONE policies on sharing data and materials.
Staphylococcus aureus is a common commensal and frequent colonizer of humans and many animal species including companion animals as well as food-producing animals. In humans, the epithelium of the anterior nares is the primary ecological niche. S. aureus is also a major pathogen involved in a wide variety of diseases such as purulent skin and subcutaneous infections, pneumonia, endocarditis, abscesses and bacteremia. Moreover, S. aureus is an emerging issue in veterinary medicine and a cause of food poisoning by its ability to produce heat-stable enterotoxins .
The transfer of S. aureus isolates between humans and animals, especially in the case of livestock-associated MRSA ST398, has recently gained particular attention . However, relatively little is known about the more global diversity of S. aureus isolates of animal origin –. This limits our ability to identify for example the origin of strains responsible for food poisoning. In order to implement control measures targeted at reservoirs and transmission routes, it is necessary to further improve current knowledge about animal-associated S. aureus.
Essentially three techniques are currently used for the large-scale analysis of the diversity of S. aureus isolates, namely multi locus sequence typing (MLST), spa typing, and multiple locus variable number of tandem repeats (VNTR) analysis (MLVA). In addition, pulsed field gel electrophoresis (PFGE) is still widely used and considered the gold-standard for typing S. aureus isolates. It has a high discriminatory power and it can be used for many bacterial pathogens. It is however not appropriate for routine interlaboratory comparisons . MLST studies allowed the description of major clonal complexes (CC) underlying the S. aureus population structure , . MLST suffers from its relatively high costs and has a moderate discriminatory power. The spa typing is a widely used method in which variations in a highly variable tandem repeat are characterized by sequencing. The Ridom Spaserver http://spaserver.ridom.de allows the designation of spa types , . The spa typing is a very powerful tool, and is currently the most commonly used first line assay. However it may fail to identify new lineages due to inherent homoplasia and variable evolutionary rate of spa alleles and clustering based on spa data is complex. MLVA was developed more recently. Homoplasia at individual VNTR loci and potentially low variability of specific alleles are compensated at least partly by the use of multiple loci. An assay comprising as little as 8 VNTR loci (called MLVA-8Bilthoven in the present report) was highly congruent with MLST and able to assign a new isolate to the correct CC for much lower costs . The 8 loci were amplified in two multiplex PCRs and analyzed by capillary electrophoresis. A MLVA assay with 14 loci (MLVA-14Orsay) providing higher discriminatory power was used in a survey of 309 isolates including clinical MRSA isolates, nasal carriage isolates and representatives of the main CCs present in humans . Both schemes can be adapted to low resolution DNA sizing equipment (such as agarose gels) as well as to higher throughput systems (such as capillary electrophoresis-based devices). MLVA data can be accessed via internet (a list of such databases is maintained on http://minisatellites.u-psud.fr). These databases can be queried even if a subset of loci is used although the discriminatory power and typing assignment precision might then be decreased.
In the present study, we have used MLVA as a first line assay, complemented when necessary by spa typing and MLST data. We have selected 16 loci for the MLVA assay, subsequently called MLVA-16Orsay, which essentially merges MLVA-8Bilthoven and MLVA-14Orsay and we have automated this assay. The products of two multiplex PCR amplifications were resolved by capillary electrophoresis, and the alleles from each of the 16 targeted loci were automatically identified. This expanded MLVA assay was used for the typing of 251 S. aureus isolates: the present retrospective investigation included 106 previously typed human clinical isolates, 98 isolates collected from various animal sources and mostly associated with a variety of diseases in these animals, 34 isolates recovered from food products, and 13 enterotoxigenic S. aureus from cases of food poisoning.
Materials and Methods
Two hundred and fifty-one isolates were included in the study. One hundred and six are human clinical isolates: ninety isolates from the HARMONY project reference collection kindly provided by Alex van Belkum were used to perform the development and initial validation of the automated MLVA protocol ; sixteen isolates were selected among two previously described collections to represent the diversity of clinical S. aureus strains from humans , . Ninety-eight isolates were previously collected from different disease conditions in farm and domestic animals , , . Thirty-four isolates were recovered from food . Thirteen isolates were associated with food poisoning (Table 1 and Table S1).
Strains were cultured overnight at 37°C in Brain Heart Infusion (BHI) or Luria Bertani broth. Genomic DNA was extracted using phenol-chloroform extraction or the DNeasy tissue kit (Qiagen, Courtaboeuf, France) after treatment with lysostaphin (100 mg/l) (Ambi products LLC, USA). Nucleic acid quality and concentration were estimated with an ND-1000 spectrophotometer (NanoDrop, Labtech, Palaiseau, France). Samples diluted in water at 5 ng/µl were used as DNA template for PCR amplification.
Selection of VNTRs and MLVA Typing
Twelve loci previously investigated by Pourcel et al. and all eight loci used by Schouls et al. ,  were merged in a single assay comprising 16 loci. The 16 VNTR loci were amplified in two multiplex PCRs using the ceeramTools® Staphylococcus typing kit (Ceeram, La Chapelle sur Erdre, France). PCR reaction 1 amplifies the ten VNTR loci Sa0122, Sa0311, Sa0387, Sa0550, Sa0684, Sa0964, Sa1097, Sa1194, Sa1729, Sa1866. PCR reaction 2 amplifies the six VNTR loci Sa0266, Sa0704, Sa1132, Sa1291, Sa2039, Sa2511 (Tables 2 and 3). VNTRs Sa0122 and Sa0266 are located in the genes spa and coa respectively. VNTRs Sa0311, Sa1729 and Sa1866 are members of the family of intergenic repeated elements called “STAR” for S. aureus repeats .
Briefly, the kit includes two primer mixes, one for each multiplex reaction. Forward primers were fluorescently labeled at the 5′ end, reverse primers were synthesized unlabeled and tailed (Applied Biosystems, Courtaboeuf, France) as previously described . Both multiplex PCRs were performed in a final volume of 15 µl using the Qiagen multiplex PCR kit (Qiagen, Courtaboeuf, France). The reactions contained 2 µl template DNA (5 ng/µl), 7.5 µl of 2× multiplex PCR mastermix and 5.5 µl of primer mix. The PCR reactions were run on a Veriti® thermal cycler (Applied Biosystems, Courtaboeuf, France) using the following conditions: initial denaturation cycle for 15 min at 95°C, 15 cycles touchdown PCR (30 s at 95°C; 60 s at 75°C, with 1.0°C drop in temperature each next cycle, 70 s at 72°C), 15 cycles long range PCR (30 s at 95°C, 60 s at 60°C, 70 s at 72°C with 5 s increase in time each next cycle), with a final 10 min step at 72°C. PCR fragments were purified using Qiagen DyeEx plates (Qiagen, Courtaboeuf, France). For each multiplex reaction, 2 µl of purified PCR product were combined with 7.75 µl HiDi formamide and 0.25 µl GS1200LIZ (Applied Biosystems, Courtaboeuf, France). Samples were loaded onto the ABI3130 capillary sequencer using a 50 cm capillary filled with performance-optimized polymer 7 (Applied Biosystems, Courtaboeuf, France) at 60°C for 6200 s with a running voltage of 12 kV, and an injection time of 10 s at an injection voltage of 1.6 kV.
MLVA Data Analysis
Each VNTR locus was identified according to color and automatically assigned a size by the GeneMapper software (Applied Biosystems, Courtaboeuf, France). This size was then converted into an allele designation associated with a quality index. Rare intermediate-sized alleles were reported as half-size (.5, Figure 1 and Figure S1) as previously described . New alleles of unexpected size were sequenced. The typing data file was imported into BioNumerics version 6.6 (Applied-Maths, Sint-Martens-Latem, Belgium). Allele designations and allele calling conventions for the VNTRs described in previous work were used as published . The MLVA code is provided in the order corresponding to the genome position in reference strain Mu50 (refseq accession number NC_002758): Sa0122, Sa0266, Sa0311, Sa0387, Sa0550, Sa0684, Sa0704, Sa0964, Sa1097, Sa1132, Sa1194, Sa1291, Sa1729, Sa1866, Sa2039, Sa2511 . Following these conventions, the genotype of the reference strain Mu50 deduced from its genomic sequence is 10-6-3-2-5-2-4-6-15-7-7-4-5-3-3-3 (Table 3).
Color coding is according to MLST clonal complex assignment whereas clustering is done according to the displayed MLVA data. Strain Id, clonal complex, sequence type, spa type, spa code and geographic origin are indicated. MLVA cluster bootstrap values are shown for the main clusters.
The diversity (D) index and confidence intervals (CI) were calculated as previously described . The UPGMA (unweighted pair group method with arithmetic mean) clustering method was run using the categorical distance coefficient. A cut-off value of 45% similarity was applied to define clusters .
Analysis of Linkage Disequilibrium, Bootstrapping and Congruence Between Different Methods
Linkage disequilibrium was measured by using LIAN Ver. 3.5 software accessed at http://guanine.evolbio.mpg.de/. The Monte-Carlo simulation was run with 100000 iterations. Bootstrap analyses were run with 500 simulations.
Comparison of Animal, Human and Food Poisoning Isolates
Fischer’s exact test was applied to compare the proportions of human-related CCs in three different populations: isolates from animals, isolates from humans and isolates involved in food poisoning. Data from the literature –,  and from the present study were analysed to build the population of animal-associated isolates. We combined our results with those described by Wattinger et al. including 20 isolates from food poisoning events, to form the population of 33 isolates involved in food poisoning .
PCR amplicons were purified using the QIAquick PCR purification (Qiagen, Courtaboeuf, France) and sequenced (Cogenics, UK or Eurofins MWG Operon, Ebersberg, Germany). Sequence data were managed with BioNumerics. The primers and conditions used for the MLST or spa tandem repeat amplification were as described by Enright et al. and Harmsen et al. respectively , . Alleles and sequence types (STs) were identified using the MLST database (http://www.mlst.net). The spa repeat nomenclature was that of Shopsin et al., and spa types were retrieved from the Ridom SpaServer http://spaserver.ridom.de .
Automated Multiplex Capillary-based MLVA Assay Development
MLVA-16Orsay integrates the two most recently published MLVA assays, each associated with large databases accessible via internet ,  (Table 2). The resulting data can be compared to both data sets. Figure 2 shows a typical capillary electrophoresis pattern of the two multiplex PCRs (Figure 2a PCR1, 10 loci; Figure 2b, PCR2, 6 loci). It preserves the convenient 2-multiplex PCR assay developed by Schouls et al. while including 8 additional VNTR loci .
a. PCR1 ten dye-colored coamplified VNTR loci. b. PCR2 six dye-colored coamplified VNTR loci.
Efficiency of the MLVA-16Orsay Protocol
The MLVA-16Orsay scheme was first tested on 90 well-described isolates of the HARMONY reference collection (Figure 1). These isolates represent epidemic or major nosocomial MRSA clones from the mid-1980s to 1998 and were previously investigated by MLVA-14Orsay . Nine missing values were observed in the present study among the expected 1440 values. Sa2511 was not amplified in five closely related isolates. Thus, MLVA-16Orsay provided high typeability (T value = 99.4%). The discriminatory index D was 0.9625, compared to 0.9531 for subset MLVA-8Bilthoven. Forty-nine types were identified when using the 16 loci, compared to 41 with the 8 loci assay.
Congruence Between MLVA-16Orsay and MLST
Direct comparison between MLST and MLVA clustering based on the 90 isolates of the HARMONY collection showed a strong correlation between these two genotyping methods (Figure 1 and Figure 3), in agreement with previous reports , . For instance, the congruence between MLST and MLVA-16Orsay was 79.3% and MLVA correctly assigned isolates to their MLST defined CC. CC nodes were supported by high bootstrap values demonstrating the reliability of MLVA-16Orsay clustering for S. aureus population investigation (Figure 1). MLVA-16Orsay was much more discriminatory than MLST (MLST distinguishes 17 STs compared to the 49 MLVA-16Orsay types and the discriminatory index D for MLST is 0.8856). The standardized index of linkage association for MLST was 0.349. In comparison the standardized index of linkage association for MLVA-16Orsay was 0.242. The detected linkage disequilibrium was highly significant in both cases (P<10−5). The lower linkage detected by MLVA was previously observed in Legionella pneumophila and might be a consequence of homoplasy at VNTR loci .
Congruence of MLVA schemes (MLVA-16Orsay, MLVA-10Orsay, and MLVA-8Bilthoven) and MLST using a Pearson correlation coefficient.
Diversity Among S. aureus Isolates Collected from Animals
The MLVA-16Orsay assay was used to type 98 animal isolates. A full data set was obtained with one exception: VNTR Sa1291 could not be amplified in isolate sa263. The 98 isolates were resolved into 59 MLVA-16Orsay types (MTs) distributed in 12 clusters and 4 singletons as shown on the minimum spanning tree of Figure 4. MLVA-8Bilthoven resolved 47 MTs (D = 0.9571). In terms of diversity, human-associated (106 isolates) and animal-associated (98 isolates) isolates were similar, with MLVA-16Orsay D values of 0.9727 and 0.9788, respectively.
Minimum spanning tree of the 251 S. aureus isolates (106 human-associated isolates, 98 animal-associated isolates and 47 isolates from food products among which 13 were related with food-poisoning) using MLVA-16Orsay. Each circle represents a MLVA genotype. The size of each circle indicates the number of isolates within this MLVA genotype. The different clusters are annotated. The host origin is indicated with a specific color. Isolates involved in food poisoning events are represented by black circles. Human and food isolates are highlighted with two different hatch patterns.
CC Distribution of the Animal-associated Isolates Deduced from Previously Published or de novo MLST Data
The tentative identification of the clusters in the animal and food-product isolates was done by comparison with previously published data , , . The spa typing was used to confirm some of the assignments. The largest clusters were CC398 and CC5, which comprised 27% (26 isolates) and 17% (17 isolates) of the typed isolates, respectively (Figure S1). Eleven MLVA-16Orsay genotypes were observed in CC398. Three MLVA-16Orsay clusters comprising 12, 9 and 2 isolates could be assigned to CC9, CC133 and CC97 respectively (Figure S1). Figure 4 shows the distribution of animal strains among the different CCs. The 32 porcine isolates clustered into 5 CCs, CC398 (12 isolates, 38%), CC9 (12 isolates, 38%), CC30 (5 isolates, 15%), CC1 (2 isolates, 6%) and CC97 (1 isolate, 3%). All porcine MRSA isolates belonged to CC398. The 30 isolates from poultry were exclusively MSSA and clustered mainly in 2 CCs: CC398 (13 isolates, 43%) and CC5 (15 isolates, 50%). The 17 isolates from dogs and cats were all MSSA, except for one MRSA CC398 isolate. They were distributed into 4 CCs, CC45 (6 isolates, 35.3%), CC8 (4 isolates, 23.5%), CC15 (3 isolates, 17.6%), CC5 (2 isolates) and a singleton (sa263). The 11 isolates from small ruminants comprised nine CC133 isolates (these MSSA isolates were obtained from sheep of the same German flock suggesting the presence of an epidemic strain) and two non-grouped isolates from mastitis. The five isolates from horses, all MSSA, belonged to three clusters, CC1 (two isolates), CC30 (two isolates) and CC25 (one isolate). Notably, both CC1 isolates shared the same MLVA-16Orsay profile although they were collected in two different countries and 16 years apart. Similarly, the two rodent isolates belonged to the same rare lineage (CC51) although they were collected in two different places and from different disease conditions.
CC Distribution Among the Food-associated Isolates of the Test Collection
Two groups of isolates were recovered from food: 13 isolates originated from cases of food-poisoning and 34 MRSA isolates originated from food not related to poisoning events (food isolates are exclusively MRSA because of the screening procedure ). Seven of the 13 food-poisoning associated isolates were collected from dairy products, five were recovered from meat products, and the precise food origin of one isolate was unknown (Figure 4, Table 1 and Table S1). Altogether, 6 isolates belonged to CC8, and single isolates to CC1, CC5, CC25, CC45, CC97, two isolates were not assigned to CCs. The second group was almost exclusively composed of CC398 isolates from poultry meat or poultry meat products (29 of 34 isolates). The remaining five isolates belonged to CC5, CC9 or CC133. Figure S2 shows a minimum spanning tree of all 55 CC398 isolates from animal or food investigated. Figure S3 shows a minimum spanning tree of all isolates based upon MLVA-16Orsay data. Animal and food samples are colored (green for MSSA samples, red for MRSA samples, poultry samples are cross-hatched).
Meta-analysis of Human-related S. aureus CCs Prevalence in Food Products Involved in Staphylococcal Intoxications
We searched the literature for reports investigating S. aureus in animals. We identified 836 farm animal isolates for which a CC assignment is known , , , , , , , . Ten (1.2%) and four (0.5%) belonged to CC8 and CC45, respectively. In contrast CC8 and CC45 represent 10–40% and 10–20% of human isolates respectively , , , , . Thirty percent (10/33) and twenty one percent (7/33) of the thirty-three isolates involved in food poisoning investigated in this study or by Wattinger et al. , belonged to CC8 and CC45 respectively. The difference between the proportion of CC8/CC45 in the animal-associated isolates population and the food poisoning isolates is highly significant according to Fisher’s exact test (P<0.05). In sharp contrast, the proportion of CC8/CC45 among food poisoning isolates is highly similar to the proportion of CC8/CC45 isolates among human isolates. This observation strongly suggested that the isolates associated with food poisoning were mainly of human origin.
High Discriminatory Power of MLVA-16Orsay in CC398
MLVA-16Orsay distinguished 19 genotypes with a diversity index of 0.8728 (Figure S2 part A). In comparison, 9 genotypes are resolved when using the MLVA-8Bilthoven subset, with a much lower diversity index of 0.6728 (Figure S2 part B). This difference is largely due to locus Sa1291 (Figure S1) which is not otherwise an exceptionally variable locus (Table 3).
The trace-back analysis of food-borne infections requires the availability of appropriate genotyping tools, i.e. highly cost-efficient and fast methods for high-throughput analysis, backed up by relevant and easily accessible typing databases. The present investigation further illustrates the relevance of MLVA for large scale population investigations of S. aureus, a major pathogen and source of food intoxications.
Automated Capillary-based 16 loci MLVA Assay
In the commercially available typing kit (ceeramTools®, Ceeram, La Chapelle sur Erdre, France), 16 loci are amplified in two multiplex PCRs. These 16 loci were chosen in order to provide data directly comparable with previously developed databases based upon the typing of 2 subsets of these loci , . The assay is able to correctly assign S. aureus isolates to defined MLST clonal complexes and further differentiate within these CCs. Homoplasy associated with VNTR loci is presumably efficiently compensated by the analysis of multiple loci. MLVA is equally well adapted for studying S. aureus epidemiology regardless of the sample origin (animal or human). The four steps of the MLVA procedure (DNA extraction, amplification of the 16 VNTRs in 2 multiplex PCRs, fragment analysis by capillary electrophoresis, MLVA code assignment) were standardized to be usable and understandable by non-expert users. The resulting data can be queried against freely accessible internet MLVA databases such as http://mlva.u-psud.fr.
MLVA-based Diversity Analysis of Animal-associated S. aureus
Ninety-six percent of the isolates were clustered within twelve known CCs. Based on MLVA results, the studied animal-associated isolates were globally as diverse as human-associated ones suggesting that, as a rule, the occurrence of S. aureus among animals is not a recent event.
Approximately half of the animal isolates belonged to eight well-known human CCs in agreement with previous studies showing the wide host range of some CCs , , . However, recent host adaptation, sporadic contamination or the presence of widespread lineages must also be taken into account as possible explanations. Evidence for recent host adaptation is provided by the well-described CC5 isolates in poultry . In the present study, companion animal-associated isolates are almost exclusively found in three main clinical-associated CCs (CC8, CC15 and CC45). This illustrates the spread of S. aureus isolates from humans to animals .
Animal-adapted Clonal Complexes
In contrast, the other half of the animal isolates is assigned to CCs not found or uncommon in humans (CC9, CC97, CC133 and CC398) . This observation confirms previous studies suggesting the existence of animal-specific lineages , , , , , , . Recently, Guinane et al. have provided evidence that CC133 which is a frequent colonizer of small ruminants, evolved as a host switch from human to ruminant followed by adaptive genome diversification . MSSA-CC9 isolates have been reported in pig farmers and also from infections of swine in France and Germany showing that CC9 is able to change hosts and colonize humans , . Two MRSA-CC9 from chicken meat were identified . This lineage is very uncommon and recently emerged from its porcine reservoir in Asia , , , . Clinical human cases due to MRSA-CC9 appeared in the same period .
The Emerging CC398
In this study, CC398 was found equally in isolates collected from pigs or poultry. Half of the poultry-associated isolates belonged to CC398 whereas CC5, rather than CC398, was previously shown to be the dominant lineage in poultry. This is, to our knowledge, the second report of MSSA CC398 isolates from poultry. This observation might suggest that the lineage was already present among poultry as MSSA and has subsequently evolved as MRSA by independent acquisition of different SCCmec elements. MSSA CC398 could have disseminated from pigs to other food-producing animals, perhaps via farm workers, and the SCCmec cassette could have been acquired in other hosts. Alternatively strains may spontaneously excise part or all of the SCCmec and thus reverse to MSSA , . Within CC398, various closely related spa types have been described (i.e. t011, t034, t108, t539 and t1793). Schouls and colleagues investigated 216 isolates belonging to CC398, among which 100 were pig-related . They observed little diversity within this complex using MLVA-8Bilthoven (D = 0.721). A similar results was obtained here with MLVA-8Bilthoven (9 genotypes in 55 isolates, D = 0.6728). In contrast, the 16-loci MLVA assay discriminates 19 genotypes (D = 0.8728), suggesting that it might be of high interest to further differentiate CC398 isolates. Due to the higher multiplexing achieved in the MLVA-16Orsay assay as described here compared to MLVA-8Bilthoven, the typing of 16 loci instead of 8 does not significantly increase the cost and workload.
MLVA Typing as A Microbial Source Tracking Tool
The present investigation suggested that S. aureus isolates involved in food poisoning are mainly strains found in humans rather than in animals. Among the 33 isolates sampled from food products involved in food poisoning and investigated in this study and by Wattinger et al. , 10 and 7 belonged to the predominantly human clusters CC8 and CC45, respectively. Given the high human host specificity of CC8 and CC45, this finding provided evidence for the role of humans as a major source of contamination. In sharp contrast, food isolates not associated with food poisoning were almost exclusively assigned to animal-specific clones (CC398, CC9 and CC133) and no CC8 or CC45 isolates were found. In the present study, the CC97 isolate 363F is the only isolate implicated in a food intoxication event with a most likely animal origin. Although these observations need to be substantiated by the analysis of a larger test population, they point towards the role of humans in the contamination of food with enterotoxigenic S. aureus.
In conclusion, we merged in this work previous MLVA schemes in a rapid and efficient automated multiplex capillary-based MLVA assay for the high-resolution genotyping of S. aureus isolates. The numeric MLVA code is produced automatically and a quality score can be defined facilitating the development of quality controlled databases. The described MLVA-16Orsay assay ensures the same clustering as MLST, assigning similarly S. aureus isolates to MLST defined clonal complexes. S. aureus MLVA typing data from the 16 loci or any convenient subset can be queried via http://mlva.u-psud.fr. MLVA is well-suited and compatible for genotyping of animal-associated S. aureus as well as human isolates. The present molecular typing analysis provided further insights into the diversity of animal-associated S. aureus. We highlighted that the animal-associated population is very diverse suggesting that animal colonization by S. aureus is globally ancient. In the present study, the S. aureus isolates from animals were divided into human-related and animal-specific CCs. Some CCs are able to switch between a great variety of hosts (i.e. CC5, CC45), whereas others seem to be strongly specific to particular human or animal hosts (i.e. CC9, CC133, CC8). The presented data indicates that S. aureus isolates from cases of food poisoning were most likely of human origin.
Dendrogram deduced from the clustering of the 251 S. aureus animal-associated isolates and human strains from the HARMONY collection using MLVA-16Orsay. The color code reflects CC assignment.
Minimum spanning tree showing the relative discriminatory power of MLVA-8Bilthoven and MLVA-16Orsay for typing CC398. The color code reflects the spa type. Part A (left): MST based upon the full MLVA-16Orsay data. Nineteen genotypes are resolved. Part B (right): MST based upon the MLVA-8Bilthoven subset of loci.
Minimum spanning tree for the poultry isolates. The minimum spanning tree is identical to the one shown in Figure 4 except for the color code. All human isolates are grayed, MSSA animal and food isolates are shown in green, MRSA isolates in red. Poultry isolates collecting from living animals are cross-hatched with horizontal lines. Poultry isolates from food products are cross-hatched with vertical lines. The MSSA CC45 poultry isolate was collected from a cooked chicken involved in a food poisoning event.
Isolates used in this study.
We thank Roswitha Becker and Vivian Hensel for excellent technical assistance. We thank Marie-Laure De Buyser from ANSES for her help at the onset of this project.
Conceived and designed the experiments: DS SS AB BL FLH CP GV. Performed the experiments: DS. Analyzed the data: DS CP GV. Contributed reagents/materials/analysis tools: DB SS AB ATF FBG KK MT. Wrote the paper: DS SS AB CP GV.
- 1. Le Loir Y, Baron F, Gautier M (2003) Staphylococcus aureus and food poisoning. Genet Mol Res 2: 63–76.
- 2. Smith TC, Pearson N (2011) The emergence of Staphylococcus aureus ST398. Vector Borne Zoonotic Dis 11: 327–339.
- 3. Hasman H, Moodley A, Guardabassi L, Stegger M, Skov RL, et al. (2010) Spa type distribution in Staphylococcus aureus originating from pigs, cattle and poultry. Vet Microbiol 141: 326–331.
- 4. Smyth DS, Feil EJ, Meaney WJ, Hartigan PJ, Tollersrud T, et al. (2009) Molecular genetic typing reveals further insights into the diversity of animal-associated Staphylococcus aureus. J Med Microbiol 58: 1343–1353.
- 5. Sung JM, Lloyd DH, Lindsay JA (2008) Staphylococcus aureus host specificity: comparative genomics of human versus animal isolates by multi-strain microarray. Microbiology 154: 1949–1959.
- 6. Rabello RF, Moreira BM, Lopes RM, Teixeira LM, Riley LW, et al. (2007) Multilocus sequence typing of Staphylococcus aureus isolates recovered from cows with mastitis in Brazilian dairy herds. J Med Microbiol 56: 1505–1511.
- 7. Aires-de-Sousa M, Parente CE, Vieira-da-Motta O, Bonna IC, Silva DA, et al. (2007) Characterization of Staphylococcus aureus isolates from buffalo, bovine, ovine, and caprine milk samples collected in Rio de Janeiro State, Brazil. Appl Environ Microbiol 73: 3845–3849.
- 8. Battisti A, Franco A, Merialdi G, Hasman H, Iurescia M, et al. (2010) Heterogeneity among methicillin-resistant Staphylococcus aureus from Italian pig finishing holdings. Vet Microbiol 142: 361–366.
- 9. Jørgensen HJ, Mork T, Caugant DA, Kearns A, Rorvik LM (2005) Genetic variation among Staphylococcus aureus strains from Norwegian bulk milk. Appl Environ Microbiol 71: 8352–8361.
- 10. Concepción Porrero M, Hasman H, Vela AI, Fernández-Garayzábal JF, Domínguez L, et al. (2012) Clonal diversity of Staphylococcus aureus originating from the small ruminants goats and sheep. Vet Microbiol 156: 157–61.
- 11. van den Berg S, van Wamel WJ, Snijders SV, Ouwerling B, de Vogel CP, et al. (2011) Rhesus macaques (Macaca mulatta) are natural hosts of specific Staphylococcus aureus lineages. PLoS One 6: e26170.
- 12. Sieber S, Gerber V, Jandova V, Rossano A, Evison JM, et al. (2011) Evolution of multidrug-resistant Staphylococcus aureus infections in horses and colonized personnel in an equine clinic between 2005 and 2010. Microb Drug Resist 17: 471–478.
- 13. de Almeida LM, de Almeida MZ, de Mendonça CL, Mamizuka EM (2011) Novel sequence types (STs) of Staphylococcus aureus isolates causing clinical and subclinical mastitis in flocks of sheep in the northeast of Brazil. J Dairy Res 78: 373–378.
- 14. Sakwinska O, Giddey M, Moreillon M, Morisset D, Waldvogel A, et al. (2011) Staphylococcus aureus host range and human-bovine host shift. Appl Environ Microbiol 77: 5908–5915.
- 15. Lin Y, Barker E, Kislow J, Kaldhone P, Stemper ME, et al. (2011) Evidence of multiple virulence subtypes in nosocomial and community-associated MRSA genotypes in companion animals from the upper midwestern and northeastern United States. Clin Med Res 9: 7–16.
- 16. Monecke S, Coombs G, Shore AC, Coleman DC, Akpaka P, et al. (2011) A field guide to pandemic, epidemic and sporadic clones of methicillin-resistant Staphylococcus aureus. PLoS One 6: e17936.
- 17. Hata E, Katsuda K, Kobayashi H, Uchida I, Tanaka K, et al. (2010) Genetic variation among Staphylococcus aureus strains from bovine milk and their relevance to methicillin-resistant isolates from humans. J Clin Microbiol 48: 2130–2139.
- 18. te Witt R, van Belkum A, MacKay WG, Wallace PS, van Leeuwen WB (2010) External quality assessment of the molecular diagnostics and genotyping of meticillin-resistant Staphylococcus aureus. Eur J Clin Microbiol Infect Dis 29: 295–300.
- 19. Feil EJ, Li BC, Aanensen DM, Hanage WP, Spratt BG (2004) eBURST: inferring patterns of evolutionary descent among clusters of related bacterial genotypes from multilocus sequence typing data. J Bacteriol 186: 1518–1530.
- 20. Robinson DA, Enright MC (2004) Multilocus sequence typing and the evolution of methicillin-resistant Staphylococcus aureus. Clin Microbiol Infect 10: 92–97.
- 21. Enright MC, Day NP, Davies CE, Peacock SJ, Spratt BG (2000) Multilocus sequence typing for characterization of methicillin-resistant and methicillin-susceptible clones of Staphylococcus aureus. J Clin Microbiol 38: 1008–1015.
- 22. Harmsen D, Claus H, Witte W, Rothganger J, Turnwald D, et al. (2003) Typing of methicillin-resistant Staphylococcus aureus in a university hospital setting by using novel software for spa repeat determination and database management. J Clin Microbiol 41: 5442–5448.
- 23. Schouls LM, Spalburg EC, van Luit M, Huijsdens XW, Pluister GN, et al. (2009) Multiple-locus variable number tandem repeat analysis of Staphylococcus aureus: comparison with pulsed-field gel electrophoresis and spa-typing. PLoS One 4: e5082.
- 24. Pourcel C, Hormigos K, Onteniente L, Sakwinska O, Deurenberg RH, et al. (2009) Improved multiple-locus variable-number tandem-repeat assay for Staphylococcus aureus genotyping, providing a highly informative technique together with strong phylogenetic value. J Clin Microbiol 47: 3121–3128.
- 25. Cookson BD, Robinson DA, Monk AB, Murchan S, Deplano A, et al. (2007) Evaluation of molecular typing methods in characterizing a European collection of epidemic methicillin-resistant Staphylococcus aureus strains: the HARMONY collection. J Clin Microbiol 45: 1830–1837.
- 26. Vu-Thien H, Hormigos K, Corbineau G, Fauroux B, Corvol H, et al. (2010) Longitudinal survey of Staphylococcus aureus in cystic fibrosis patients using a multiple-locus variable-number of tandem-repeats analysis method. BMC Microbiol 10: 24.
- 27. Schwarz S, Kadlec K, Strommenger B (2008) Methicillin-resistant Staphylococcus aureus and Staphylococcus pseudintermedius detected in the BfT-GermVet monitoring programme 2004–2006 in Germany. J Antimicrob Chemother 61: 282–285.
- 28. Kadlec K, Ehricht R, Monecke S, Steinacker U, Kaspar H, et al. (2009) Diversity of antimicrobial resistance pheno- and genotypes of methicillin-resistant Staphylococcus aureus ST398 from diseased swine. J Antimicrob Chemother 64: 1156–1164.
- 29. Strommenger B, Kehrenberg C, Kettlitz C, Cuny C, Verspohl J, et al. (2006) Molecular characterization of methicillin-resistant Staphylococcus aureus strains from pet animals and their relationship to human isolates. J Antimicrob Chemother 57: 461–465.
- 30. Feßler AT, Kadlec K, Hassel M, Hauschild T, Eidam C, et al. (2011) Characterization of methicillin-resistant Staphylococcus aureus isolates from food and food products of poultry origin in Germany. Appl Environ Microbiol 77: 7151–7157.
- 31. Cramton SE, Schnell NF, Gotz F, Bruckner R (2000) Identification of a new repetitive element in Staphylococcus aureus. Infect Immun 68: 2344–2348.
- 32. Sobral D, Le Cann P, Gerard A, Jarraud S, Lebeau B, et al. (2011) High-throughput typing method to identify a non-outbreak-involved Legionella pneumophila strain colonizing the entire water supply system in the town of Rennes, France. Appl Environ Microbiol 77: 6899–6907.
- 33. Kuroda M, Ohta T, Uchiyama I, Baba T, Yuzawa H, et al. (2001) Whole genome sequencing of meticillin-resistant Staphylococcus aureus. Lancet 357: 1225–1240.
- 34. Haubold B, Hudson RR (2000) LIAN 3.0: detecting linkage disequilibrium in multilocus data. Linkage Analysis. Bioinformatics 16: 847–848.
- 35. Wattinger L, Stephan R, Layer F, Johler S (2012) Comparison of Staphylococcus aureus isolates associated with food intoxication with isolates from human nasal carriers and human infections. Eur J Clin Microbiol Infect Dis 31: 455–64.
- 36. Shopsin B, Gomez M, Montgomery SO, Smith DH, Waddington M, et al. (1999) Evaluation of protein A gene polymorphic region DNA sequencing for typing of Staphylococcus aureus strains. J Clin Microbiol 37: 3556–3563.
- 37. Visca P, D’Arezzo S, Ramisse F, Gelfand Y, Benson G, et al. (2011) Investigation of the population structure of Legionella pneumophila by analysis of tandem repeat copy number and internal sequence variation. Microbiology 157: 2582–2594.
- 38. Rijnders MI, Deurenberg RH, Boumans ML, Hoogkamp-Korstanje JA, Beisser PS, et al. (2009) Population structure of Staphylococcus aureus strains isolated from intensive care unit patients in the netherlands over an 11-year period (1996 to 2006). J Clin Microbiol 47: 4090–4095.
- 39. Argudín MA, Mendoza MC, Vazquez F, Guerra B, Rodicio MR (2011) Molecular typing of Staphylococcus aureus bloodstream isolates from geriatric patients attending a long-term care Spanish hospital. J Med Microbiol 60: 172–179.
- 40. Vainio A, Koskela S, Virolainen A, Vuopio J, Salmenlinna S (2011) Adapting spa typing for national laboratory-based surveillance of methicillin-resistant Staphylococcus aureus. Eur J Clin Microbiol Infect Dis 30: 789–797.
- 41. Lowder BV, Guinane CM, Ben Zakour NL, Weinert LA, Conway-Morris A, et al. (2009) Recent human-to-poultry host jump, adaptation, and pandemic spread of Staphylococcus aureus. Proc Natl Acad Sci U S A 106: 19545–19550.
- 42. Rich M (2005) Staphylococci in animals: prevalence, identification and antimicrobial susceptibility, with an emphasis on methicillin-resistant Staphylococcus aureus. Br J Biomed Sci 62: 98–105.
- 43. Guinane CM, Ben Zakour NL, Tormo-Mas MA, Weinert LA, Lowder BV, et al. (2010) Evolutionary genomics of Staphylococcus aureus reveals insights into the origin and molecular basis of ruminant host adaptation. Genome Biol Evol 2: 454–466.
- 44. Armand-Lefevre L, Ruimy R, Andremont A (2005) Clonal comparison of Staphylococcus aureus isolates from healthy pig farmers, human controls, and pigs. Emerg Infect Dis 11: 711–714.
- 45. Kehrenberg C, Cuny C, Strommenger B, Schwarz S, Witte W (2009) Methicillin-resistant and -susceptible Staphylococcus aureus strains of clonal lineages ST398 and ST9 from swine carry the multidrug resistance gene cfr. Antimicrob Agents Chemother 53: 779–781.
- 46. Guardabassi L, O’Donoghue M, Moodley A, Ho J, Boost M (2009) Novel lineage of methicillin-resistant Staphylococcus aureus, Hong Kong. Emerg Infect Dis 15: 1998–2000.
- 47. Wagenaar JA, Yue H, Pritchard J, Broekhuizen-Stins M, Huijsdens X, et al. (2009) Unexpected sequence types in livestock associated methicillin-resistant Staphylococcus aureus (MRSA): MRSA ST9 and a single locus variant of ST9 in pig farming in China. Vet Microbiol 139: 405–409.
- 48. Liu Y, Wang H, Du N, Shen E, Chen H, et al. (2009) Molecular evidence for spread of two major methicillin-resistant Staphylococcus aureus clones with a unique geographic distribution in Chinese hospitals. Antimicrob Agents Chemother 53: 512–518.
- 49. Neela V, Mohd Zafrul A, Mariana NS, van Belkum A, Liew YK, et al. (2009) Prevalence of ST9 methicillin-resistant Staphylococcus aureus among pigs and pig handlers in Malaysia. J Clin Microbiol 47: 4138–4140.
- 50. Chlebowicz MA, Nganou K, Kozytska S, Arends JP, Engelmann S, et al. (2010) Recombination between ccrC genes in a type V (5C2&5) staphylococcal cassette chromosome mec (SCCmec) of Staphylococcus aureus ST398 leads to conversion from methicillin resistance to methicillin susceptibility in vivo. Antimicrob Agents Chemother 54: 783–791.
- 51. Boundy S, Zhao Q, Fairbanks C, Folgosa L, Climo M, et al. (2012) Spontaneous SCCmec excision in Staphylococcus aureus nasal carriers. J Clin Microbiol 50: 469–471.
- 52. Hardy KJ, Oppenheim BA, Gossain S, Gao F, Hawkey PM (2006) Use of variations in staphylococcal interspersed repeat units for molecular typing of methicillin-resistant Staphylococcus aureus strains. J Clin Microbiol 44: 271–273.
- 53. Gilbert FB, Fromageau A, Gelineau L, Poutrel B (2006) Differentiation of bovine Staphylococcus aureus isolates by use of polymorphic tandem repeat typing. Vet Microbiol 117: 297–303.
- 54. Sabat A, Krzyszton-Russjan J, Strzalka W, Filipek R, Kosowska K, et al. (2003) New method for typing Staphylococcus aureus strains: multiple-locus variable-number tandem repeat analysis of polymorphism and genetic relationships of clinical isolates. J Clin Microbiol 41: 1801–1804.