International Journal of Food Microbiology 139 (2010) 116–125
Contents lists available at ScienceDirect
International Journal of Food Microbiology
j ourna l homepage: www.e lsev ie r.com/ locate / i j foodmicro
Short Communication
Genotypic and phenotypic characterisation of a collection of Cronobacter(Enterobacter sakazakii) isolates
Rabeb Miled-Bennour a, Timothy C. Ells b,⁎, Franco J. Pagotto c, Jeffrey M. Farber c, Annaelle Kérouanton a,Thomas Meheut a, Pierre Colin d, Han Joosten e, Alexandre Leclercq f, Nathalie Gnanou Besse a
a Agence française de sécurité sanitaire des aliments, Afssa Laboratoire d'Etudes et de Recherches sur la Qualité des Aliments et des Procédés agro-alimentaires (Afssa LERQAP),23 Avenue du Général de Gaulle, 94706 Maisons Alfort cedex, Franceb Atlantic Food and Horticulture Research Centre, Agriculture and Agri-Food Canada, 32 Main Street, Kentville, Nova Scotia, Canada B4N1J5c Listeriosis Reference Service, Bureau of Microbial Hazards, Health Canada, 251 Sir F.G. Banting Driveway, Ottawa, Ontario, Canada K1A0K9d Ecole Supérieure de Microbiologie et de Sécurité Alimentaire (ESMISAB), Plouzane, Francee Quality and Safety Department, Nestlé Research Centre, Vers-chez-les-Blanc, CH-1000 Lausanne, Switzerlandf Institut Pasteur, 25 rue du Docteur Roux, Paris, France
Enterobacter sakazakii has been identified as the causative agent of serious neonatal infections, associatedwith high mortality rate. In many cases, powdered infant formula (PIF) has been identified as the source ofinfection. Recently, E. sakazakii was proposed to be classified in a new genus, Cronobacter. Since knowledgeon this pathogen is still incomplete, there is a need for molecular characterization schemes in order to helpwith epidemiological investigation and evaluate strain variability. The objectives of this study were tocombine genotypic (pulsed-field gel electrophoresis [PFGE], 16S rRNA gene sequencing, and automatedribotyping) methods with traditional phenotypic biochemical methods to characterize a collection ofCronobacter isolates from various origins. In addition, the relative growth dynamics were compared byestimating the growth rates for each isolate in non-selective broth (BHI) at 25 °C and 37 °C. According tobiochemical test profiles the majority of isolates were identified as Cronobacter sakazakii, which seemed to bethe most common species distributed in the environment of PIF production plants. Furthermore, the PFGEtechnique displayed very high discriminatory power as 61 distinct pulsotypes were revealed among the 150Cronobacter isolates. Combining information on sample origin and pulse type, 64 isolates were deemed asunique strains. Although genetic typing data for the strains clearly delineated them into clusters closelycorresponding to biochemical speciation results, it was not without discrepancies as some strains did notgroup as predicted. Important for quantitative risk assessment is the fact that despite the high geneticheterogeneity observed for this collection, most Cronobacter strains displayed similar growth ratesirrespective of species designation.
Enterobacter sakazakii, a Gram-negative motile rod belonging tothe Enterobacteriaceae family was formerly known as a “yellowpigmented Enterobacter cloacae”. It was initially designated as aunique species in 1980 (Farmer et al., 1980), and has subsequentlybeen investigated (Iversen et al., 2007; 2008) to clarify its taxonomy.The proposed reclassification of this organism into a new genus,Cronobacter, was based on the results of independent molecularmethods and of biochemical markers (Iversen et al., 2007; 2008).
Cronobacter species are considered as emerging foodborne patho-gens, andhavebeen identifiedas the causative agent of several outbreaks
or sporadic cases of very serious neonatal infections causing meningitis,septicaemia or necrotising enterocolitis in infants (Arseni et al., 1987;Nazarowec-White and Farber, 1997a; Bar-Oz et al., 2001). The diseasefrequency is very low (only 76 cases were reported from 1961 to 2003)(Iversen and Forsythe, 2003; Gurtler et al., 2005), but the mortality ratehas been reported to be as high as 20 to 50% (Anonymous, 2004; Lehnerand Stephan, 2004) with surviving patients often suffering severeneurological sequelae (Biering et al., 1989; Anonymous, 2006a). In mostcases, the source of infection was powdered infant formula (PIF)(Simmons et al., 1989; Clark et al., 1990; Van Acker et al., 2001; Lehnerand Stephan, 2004; Anonymous, 2005). Moreover, the high resistance ofC. sakazakii to osmotic stress (including drying), contributes to itspersistence in PIF factories, products and environments (Breeuwer et al.,2003). C. sakazakii has been reported to be prevalent in 40% of dryenvironmental samples from PIF factories (Guillaume-Gentil et al.,2005). Like many other enterobacteriaceae, C. sakazakii has also been
In order to facilitate epidemiological investigations, it is recom-mended that all Cronobacter isolates should be characterized bothgenotypically and phenotypically (Gurtler et al., 2005). In recentyears, the quantitative risk assessment (QRA) has become a usefultool for improvement of the sanitary quality of food. Therefore, theacquisition of data concerning strain diversity among pathogens isessential. However, there is a general lack of information regardingdifferences between Cronobacter species (Anonymous 2004, 2006a).
Several methods may be used to study bacterial biodiversity orsource tracking. In particular, molecular based techniques havebecome important tools for the subtyping of bacteria. Among these,ribotyping, random amplification of polymorphic DNA, pulsed-fieldgel electrophoresis (PFGE), multiple-locus variable-number tandem-repeat analysis (MLVA) and repetitive sequence-based polymerasechain reaction (REP-PCR) have been successfully applied to thecharacterization of Cronobacter species (Nazarowec-White andFarber, 1999; Block et al., 2002; Drudy et al., 2006; Mullane et al.,2007; Proudy et al., 2008a, 2008b; El-Sharoud et al., 2008; Healy et al.,2008). Although a PFGE standardised protocol has not yet beenvalidated for this pathogen, this method is considered the “goldstandard” method for subtyping of foodborne bacteria, and the mostdiscriminatory technique for genetic typing (Nazarowec-White andFarber, 1999; Healy et al., 2008).
The objectives of this study were to apply genotypic andphenotypic methodologies to characterize a collection of Cronobacterstrains from various origins including, PIF factories, PIF products andenvironments, as well as clinical samples. Our investigations includedphenotypic characterization using biochemical tests and geneticprofiling using PFGE, automated ribotyping and 16S rRNA genesequencing. Additionally, in order to compare growth characteristics,growth rates were estimated for each strain in non-selective broth at25 °C and 37 °C.
2. Materials and methods
2.1. Collection of Cronobacter isolates
A collection of 150 isolates, originating from clinical samples, PIFproducts and environmental samples from PIF production plants(several European factories from France, Germany and Switzerland)were examined in this study. Of these 150 isolates, 70% were isolatedfrom the PIF production environment, 20% PIF, and 10%, includingclinical isolates, were received from the collections of other institutes.Isolates were previously identified as E. sakazakii using the ISO/TS22964 standard method for the detection of E. sakazakii in PIF(Anonymous, 2006b; Gnanou Besse et al., 2006). Briefly, themethod isbased on pre-enrichment in buffered peptone water, followed by aselective enrichment procedure in modified lauryl sulfate tryptosebroth and plating on the chromogenic selective isolation agar“Enterobacter sakazakii Isolation Agar” (ESIA). Typical colonies chosenwere those with glucosidase activity (blue on chromogenic medium)and either white or yellow pigment on TSA. Presumptive isolates werefurther confirmed using ID 32E version 3.0 biochemical galleries(bioMérieux, Marcy l'étoile, France). These confirmation tests wererepeated twice on each isolate, following a few months time interval.Stock cultures were maintained frozen at −80 °C using Cryobanktubes (AES Laboratoires, Combourg, France). Cultures were revived byplating onto tryptone soya agar with yeast extract (TSA-YE) beforeuse.
2.2. Pulsed field gel electrophoresis (PFGE) typing
In order to determine the genetic diversity within our collection,all Cronobacter isolates were subjected to PFGE, the current “gold
standard” for molecular typing of foodborne bacterial pathogens. Inthis study, isolates with different PFGE profiles, or isolates withidentical PFGE patterns but from a different origin, were considered tobe different strains. The PulseNet standardised PFGE protocol for thesubtyping of Salmonella spp. was adapted to Cronobacter (Hunter etal., 2005; Ribot et al., 2006), by performing the following modifica-tions: the quantity of restriction enzyme XbaI (Roche Diagnostics,Mannheim, Germany) used to digest genomic DNA was increased 1.5fold compared to the current practice in our laboratory (from 50 to75 U per sample). For strains known to be affected by DNAdegradation (difficulties encountered for 11% of our isolates), theTris was replaced by HEPES (N′-2-Hydroxyethylpiperazine-N′-2ethanesulphonic acid) at the same concentration in all solutions;thiourea was added to the electrophoresis buffer to a finalconcentration of 100 µM (Liesegang and Tschäpe, 2002; Koort et al.,2002; Silbert et al., 2003). Salmonella serotype Braenderup (strainH9812) was used as the DNA size marker (Hunter et al., 2005) forcomparative analysis. Comparisons were realised using Bionumerics4.5 software (Applied Maths, Sint-Martens-Latem, Belgium). Adendrogram was obtained using the unweighted pair group methodwith arithmetic mean (UPGMA) and the DICE coefficient with 1%tolerance.
2.3. Biochemical differentiation of Cronobacter species
Iversen et al. (2007), 2008 demonstrated that the importantbiochemical tests for the differentiation of Cronobacter spp. wereindole production, malonate utilization and acid production fromdulcitol and methyl-α-D-glucoside. These relevant biochemical testswere applied to all different strains (as determined by PFGE typing) inorder to classify them presumptively into Cronobacter species. Briefly,to test the acid production from carbohydrate (dulcitol andmethyl-α-D-glucopyranoside), strains were incubated in phenol red broth(containing per litre of deionized water: 10 g peptone, 1 g yeastextract, 5 g NaCl, and 0.018 g phenol red) with addition of filter-sterilized carbohydrate solution (final concentration: 0.5%). Sodiummalonate broth (Difco) was used to determine malonate utilization.Indole production was measured by the addition of James reagent(bioMérieux), after growth in peptone water without indole (bio-Mérieux). All broths were incubated for 24 h at 37 °C.
2.4. Automated ribotyping
Ribotyping was performed with the restriction enzyme EcoRI andthe RiboPrinter microbial characterization system (Qualicon Inc.,Wilmington, Del.), according to the manufacturer's manual and aspreviously described (Bruce et al., 1995; Bruce, 1996). A UPGMAdendrogram was created by downloading the riboprints to Bionu-merics (version 5.0) using a DICE coefficient with optimization of 1%and a position tolerance of 1.5%.
2.5. 16S rRNA gene sequencing
Total genomic DNAwas extracted from overnight Luria Betani (LB)broth cultures grown at 37 °C using a commercial DNA extraction kit(Mo Bio Laboratories, Carlsbad, CA, USA). Near full-length (∼1.5 kb)16S rRNA gene amplicons were generated by PCR in a T-gradientthermocycler (Biometra, Göttingen, Germany) using Taq DNApolymerase (Sigma-Aldrich Canada, Oakville, ON, Canada) and theprimer pair, SDBACTF (5′-GAG TTT GAT CMT GCG TCA G-3′) andBAC1492R (5′-TAC GGY TAC CTT GTT ACG ACT T-3′) (Lane, 1985).Cycle conditions were as follows: initial denaturation at 95 °C for2 min, followed by 30 cycles of denaturation at 94 °C for 30 s,annealing at 55 °C for 30 s, and 2 min elongation at 72 °C. A finalextension step of 5 min at 72 °C followed the final cycle. PCR productswere diluted 100 fold in TE buffer and then cloned into E. coli using a
118 R. Miled-Bennour et al. / International Journal of Food Microbiology 139 (2010) 116–125
StrataClone PCR cloning kit according to the manufacturer's instruc-tions (Stratagene, La Jolla, CA, USA).
DNA sequencing reactions were carried out using the BigDyeTerminator v3.1 cycle sequencing kit and run on an ABI 3130 geneticanalyzer (Applied Biosystems, Foster City, CA) using a 36 cm capillarycolumn containing POP7 polymer. M13 forward and reverse primerswere used to sequence the cloned amplicons from both directionsand the universal primer, 16SInternal (5′-TCA CRR CAC GAG CTG ACGA-3′), was incorporated to sequence the internal regions. Sequencerv4.5 (Gene Codes Corp, Ann Arbor, MI) was used to align fragmentsand sequences were aligned using ARB (Ludwig et al., 2004) andoperational taxonomic units (OTU) using a 97% cut off were identifiedfrom a distance matrix generated by the ARB software. ARB softwarewas used to generate an aligned sequences file and phylogenetic treeswere created using PHYLO_WIN (Galtier et al., 1996). Twenty oneadditional reference sequences including both Cronobacter spp. aswell as other closely related members of the Enterobacteriaceae werealso included in the database. Settings for each tree included a NearestNeighbour Algorithm with a Jukes–Cantor correction and pair-wisegap removal. To statistically evaluate the trees, bootstrap values werecalculated using 2000 tree iterations.
2.6. Growth rates comparison
Strains were propagated twice (respectively, 6 and 18 h cultures at37 °C) in Brain Heart Infusion broth (BHI) before use. The final BHIculture was in stationary phase and contained approximately1×109 CFU/ml. All dilutions were performed in BHI broth. Growthcurves for each strain were determined in triplicate in BHI broth, at25 °C and 37 °C, bymeasuring optical density (OD) at 600 nm using anautomated spectrophotometer (Bioscreen C reader, Labsystems,France). The initial level of inoculum in the growth medium (BHI)was approximately 1×105 CFU/ml. Growth rates (µ) were estimatedfrom the slope of the tangential line of Ln(OD) evolution in mid-exponential phase, and compared by a single factor analysis ofvariance (ANOVA). Though µ measurement through OD evolutiondoes not ascertain the exact growth rates, it is convenient for itsestimation and inter-strains comparison purposes.
3. Results and discussion
Insight into the genetic and physiological variability of pathogensis a valuable prerequisite for improving the reliability of quantitativerisk assessments. The acquisition of data concerning the diversity ofstrains of foodborne pathogens is an essential part of this framework.It has been proposed that E. sakazakii should be reclassified as fivespecies and one genomospecies belonging to the genus Cronobacter.Recently, Iversen et al. (2007), 2008 applied ribotyping, amplifiedfragment length polymorphisms (AFLP) and 16S rRNA gene sequenc-ing in combination with DNA–DNA hybridizations to a collection of210 E. sakazakii isolates. Their results not only formed the basis for thereclassification of this pathogen into the new Cronobacter genus butalso clarified species designations. Based on the new nomenclaturewe have characterized a collection of isolates of Cronobacter spp.
For our collection, a highdegree of genetic diversitywas revealed asPFGE after genomic DNA restriction using XbaI resulted in 61 uniqueprofiles (i.e., clustering by genetic relatedness between isolateswas not observed at more than 60% of similarity) (Fig. 1). Moreover,isolates originating from the same processing facilities also varied inPFGE profiles, a trend that has been reported in other studies (Gadzovet al., 2006; Proudy et al., 2008b; Mullane et al., 2007; 2008b; Healy etal., 2008). This illustrates the variety of ecological niches and possiblesources of contamination that could impact the sanitary quality of endproducts leaving PIF production facilities.
In order to declare isolates as individual strains our criteria entailedthat; 1) isolatesmust have aunique PFGEprofile, or 2) isolateswith the
same profile must have different sample origins. Based on theserequirements we designated 64 isolates as specific strains. Of thisgroup, 66% had been isolated from PIF production environments, 8%from PIF, 3% from other foods, and 9% were of clinical origin.Information pertaining to sample origin could not be obtained for14% of these strains. However, no relationship between pulsotype andsample type could be ascertained. The 64 strains were furthersubjected to discriminatory biochemical tests conducted in twoindependent laboratories. The results indicated that 82.5% of thestrains were C. sakazakii, 8% C. malonaticus, 5% C. muyjtensii, 3% C.dublinensis, and 1.5% C. turicensis (Fig. 1, Table 1). Of the 47 strainsisolated from either PIF products or environmental samples from PIFproduction plants, 91.5% were C. sakazakii, and 8.5% were C.malonaticus indicating the prevalence of C. sakazakii distributed inPIF factories, products and environments. A previous study on acollection of Cronobacter isolates from various origins also displayed ahigher number of C. sakazakii and C. malonaticus strains from PIFproducts and production environments (Healy et al., 2008). Mullaneet al. (2008a) also reported similar results for samples obtained from aunique powdered milk-protein factory.
The phylogenetic relatedness of our isolateswas investigated usinga combination of 16S rRNA gene sequencing and automated ribotyp-ing. Ribotyping results obtained here clearly defined groupings whichcorresponded to their speciation according to biochemical analysisconducted on the strains (Fig. 2). However, two notable exceptionswere 05CHPL02 (CDC 28-83) and 05CHPL53. Although the formerisolate was identified as C. sakazakii through biochemical tests,ribotyping placed it closer to the non-sakazakii strains in the collection.Conversely, 05CHPL53 was biochemically identified as C. malonaticusbut ribotyping categorized it as C. sakazakii. These tests were repeatedand performed in two independent labs to confirm these results.
Sequencing of the 16S rRNA gene followed by phylogeneticanalysis provided a level of discretion which separated C. sakazakiifrom other species within the genus (Fig. 3). In order to declare twoisolates as separate species by 16S rRNA gene sequencing, it isnecessary to sequence the entire 16S rRNA gene and the similaritybetween the two must be b98.3–99% (Stackebrandt and Ebers, 2006).Therefore ∼1.5 kb of the 16S rRNA gene was sequenced here and theGenBank accession numbers are given in Table 2. Within the C.sakazakii grouping there also appeared to be three sub-groups; withsubgroup 1 containing 16 isolates including ATCC BAA894; subgroup2 containing 19 members with reference strain E266 and; subgroup 3having 22 isolates including reference strains ATCC 29544 and JMC1233 (Fig. 3). However, attempts to further categorize these isolates atthis level may be somewhat artificial since there is N99.3% similarity inthe 16S rRNA gene sequence between the most distal members ofsubgroup 1 and 3. Further discretion could be achieved by employinga DNA–DNA hybridization assay (Iversen et al., 2007).
Although most isolates phenotypically designated as C. sakazakiigrouped into a defined cluster (three sub groupings), the otherspecies did not fall completely into the other three clusters as wewould have predicted according to results from our ribotyping andbiochemical analysis. Nor were these categorizations as seamless asthose described in the Iversen et al. study (2007). For example strains05CHPL40 and 05CHPL53, phenotypically identified as C. malonaticus,clustered with the second C. sakazakii sub-group, and two otherisolates of the C. malonaticus phenotype (05CHPL46, 07HMPA87A)appeared to be more closely related to the C. sakazakii group ratherthan forming their own cluster. Furthermore, collection strain05CHPL02 (CDC 28-83) phenotypically characterized as C. sakazakii,also clustered outside its expected group. According to ribotypingresults, 4 of these 5 strains were also separated from the C. sakazakiigroup, the exception being 05CHPL53 (Fig. 2). Fractionation of theother Cronobacter species into distinct clades by 16S rRNA sequencingalso was not as clear. For example, C. muytjensii strains formed theirown clade but strains identified as C. turicensis, C. dublinensis and C.
Fig. 1. Dendrogram showing XbaI-mediated pulsed-field gel electrophoresis (PFGE) profiles of Cronobacter strains. The dendrogram was obtained using the unweighted pair groupmethod with arithmetic mean (UPGMA) and the DICE coefficient with 1% tolerance.
119R. Miled-Bennour et al. / International Journal of Food Microbiology 139 (2010) 116–125
Table 1List of Cronobacter strains used in this study along with their PFGE type, origin andgrowth rates in BHI at 25 °C and 37 °C.
Species PFGEtype
Strain Origin Growthrates (h−1)
25 °C 37 °C
C. sakazakii ESXB001 05CHPL01 Clinical (type strainATCC 29544); child'sthroat
C. muytjensii ESXB023 05CHPL63 Food 0.33 0.60ESXB032 05CHPL61 Collection strain 0.35 0.60ESXB036 05CHPL67 Type strain (ATCC
51329)0.35 0.63
C. dublinensis ESXB035 05CHPL64 Rodent food (collectionstrain)
0.35 0.65
ESXB053 08HMPA03 Type strain (DSMZ18705)
0.32 0.59
C. turicensis ESXB060 08HMPA02 Clinical (type strainDSMZ 18703); neonatalmeningitis
0.32 0.65
120 R. Miled-Bennour et al. / International Journal of Food Microbiology 139 (2010) 116–125
malonaticus formed another group. There were also deviations fromthis pattern as strain 08HMPA03 (DSMZ 18705), identified as C.dublinensis according to ribotyping and phenotypic tests, clusteredwith the C. muytjensii group. However, according to the ribotypingresults, strains that were phenotypically identified as C. dublinensis orC. muytjensii were more closely related to each other, whereas the C.malonaticus/C. turicensis group demonstrated closer similarity to C.sakazakii (Fig. 2). Moreover, strain 05CHPL65, identified as C. sakazakiiaccording to phenotype, clustered with the mixed group according to16S rRNA gene sequence. Ribotyping corroborated these findings asthis strain branched from the same point as 08HMPA01, another strainidentified as C. malonaticus. Irrespective of grouping it is important tonote that, according to 16S rRNA gene sequence, the two most distalstrains in our collection (05CHPL64 and 05CHPL50) still shared greaterthan 97% sequence identity. Another interesting observation from ourphylogenetic tree was that other closely related species within theEnterobacteriacae, such as Escherichia coli displayed greater sequencesimilarity to Cronobacter spp. than did strains representative ofEnterobacter cloacae (Fig. 3). For example, reference strain C. sakazakiiATCC 29544 had 96% sequence similarity with E. coli ATCC 29522,while only 94% similarity with the E. cloacae ATCC 13047T.
Due to the limited resolution of the 16S rRNA gene at the specieslevel, discrepancies in sequencing data are not uncommon (Konstan-tinidis and Tiedje, 2007). Other methods may be better suited forphylogenetic studies of Cronobacter. Recently, greater discretion forthe speciation of Cronobacter spp. has been demonstrated usingmultilocus sequence analysis (MLST) (Baldwin et al. 2009; Kuhnert etal., 2009). In particular, sequencing of recN alone appears to provide ahigher degree of utility over the 16S rRNA gene (El-Sharoud et al.,2009). In our study, 16S rRNA gene sequencing revealed that strainsgrouped in any of the three clusters outside of the C. sakazakii grouphad at least 98% sequence similaritywith one another, thus illustratingthe difficulty in accurate speciation using this method exclusively.However, the relatedness of individual strains within these genotypeclusters can be further elucidated by conducting the appropriateregimen of biochemical tests. Farmer et al. (1980) described 15biogroups for E. sakazakii based on 10 defining biochemical tests.Iversen et al. (2006) later demonstrated that these specific biogroupscould be placed within specific genotype clusters based on partial 16SrRNA gene sequencing. It should also be noted that no relationshipcould be ascertained from the results in our study or previous work(Iversen et al., 2006) with respect to sample origin as strains from PIF,the processing environment or clinical samples did not form sub-clusters irrespective of the typing method utilized.
Information regarding strain diversity, aswell as the dynamics of thegrowth of these pathogens is essential for the mitigation of risk sincedifferent bacterial species may display different behaviours. Therefore,we determined the growth rates for each of our strains in BHI broth at25 °C and 37 °C (Table 1). These temperatures were chosen in order tomimic conditions after reconstitution of PIF in feedingbottles at ambient
Fig. 2. Automated ribotyping analysis of a collection of Cronobacter strains. Dendrogramwas generated using a UPGMA algorithm and a DICE coefficient with optimization of 1% and aposition tolerance of 1.5%. Species designations are given according to biochemical profiles. *05CHPL02was identified as C. sakazakii by biochemical tests but grouped outside speciescluster. **05CHPL53 was identified as C. malonaticus by biochemical tests but grouped with C. sakazakii.
121R. Miled-Bennour et al. / International Journal of Food Microbiology 139 (2010) 116–125
Fig. 3. Phylogenetic tree for a collection of Cronobacter strains and other closely related Enterobacteraceae based on 16S rRNA gene sequencing. A Nearest Neighbour algorithm wasimplemented with a Jukes–Cantor correction and pair-wise gap removal. Bootstrap values were calculated using 2000 tree iterations.
122 R. Miled-Bennour et al. / International Journal of Food Microbiology 139 (2010) 116–125
temperature or after warming, respectively. Growth rates ranged from0.20 to 0.38/h at 25 °C and from 0.38 to 0.80/h at 37 °C (Table 1).Analysis of variance (ANOVA, single factor) revealed that although there
were significant differences (Pb0.05) in growth rates among individualstrains, there was no significant difference (PN0.05) in mean growthrate between the origin of the strain or between species (Table 1, Fig. 4).
Table 2GenBank accession numbers for the 16S rRNA gene.
Fig. 4. Comparison of the mean growth rate (μ) for Cronobacter strains in BHI at 25 °C( ) and 37 °C ( ), according to A.) species or B.) origin (n=number of tested strains).Growth rates of individual strains are presented in Table 1.
123R. Miled-Bennour et al. / International Journal of Food Microbiology 139 (2010) 116–125
Moreover, no links between strain origin, genotype or phenotype(growth rate) could be established.
Despite the high level of genetic heterogeneity between strainsthere was a high level of homogeneity in growth behaviour in BHIbetween both species and strain origin. Low intraspecific growthvariability in broth at optimal or suboptimal temperature has alsobeen observed for several other bacteria, including, Listeria mono-cytogenes, Escherichia coli, Clostridium perfringens and Bacillus cereus(Membré et al., 2005). In some studies, E. sakazakii intraspecificgrowth variability has been estimated, but for fewer strains(Nazarowec-White and Farber, 1997b; Lenati et al., 2008). In contrastto our results, Cooney et al. (2009) observed significant differences ingrowth rates at 37 °C in reconstituted PIF for Cronobacter species, butthis trend may have been biased by the low number of strains used inthe study. Also, Lenati et al. (2008), using nine E. sakazakii isolates,observed that the average generation time of clinical strains waslonger at 37 °C, as compared to environmental and food isolates.Contrary to this, Nazarowec-White and Farber (1997b) did notobserve significant differences between generation times for clinicalor food isolates at 23 °C in infant formula. Certainly, growth rates willbe impacted by temperature and choice of medium as Iversen et al.(2004) observed a broader doubling time at 37 °C for six strains of E.sakazakii using BHI rather than PIF.
The relationship between growth rates in Cronobacter species,genotype or origin could be better evaluated by examining a muchbroader and diverse collection of isolates. The mean generation timefor our isolates grown in BHI at 37 °C was approximately 1.1 h, whichis quite similar to results found in other studies examining E. sakazakiiin rich media, such as BHI or PIF at 37 °C (Iversen et al., 2004; Kandhaiet al., 2006; Lenati et al., 2008; Cooney et al., 2009). At 25 °C, weobtained a mean generation time of 2.1 h. Previous work conductedusing infant formula at similar temperatures demonstrates thevariability in mean generation times (Nazarowec-White and Farber,1997b; Iversen et al., 2004). Although no correlation between isolate
origin and growth rates could be drawn from our results weacknowledge that our collection was dominated with environmentalstrains. The inclusion of a greater number of clinical isolates would berequired to provide a better assessment of possible relationships.
Another important growth parameter is the lag phase duration.Kandhai et al. (2006) examined effects of pre-culturing conditions onlag times of E. sakazakii in PIF. In order to elucidate the true behaviourof foodborne pathogens within food samples, future studies should beperformed on the lag phase distribution among the isolates usingstressed cells in relevant media such as reconstituted infant formulasince the microbiological quality of PIF is in question (Muytjens et al.,1988). Extreme temperature conditions, closer to the growth limits,are also likely to reveal differences between strains for bacterialgrowth (Barbosa et al., 1994; Lebert et al., 1998; Bergis et al., 2004;Rosset et al., 2007).
Using a combination of biochemical and molecular techniques,we have characterized a collection of isolates previously identifiedas E. sakazakii. Results obtained here allowed for their designationinto species within the new Cronobacter genus and the identificationof unique strains. However, relationships between strain originsor growth dynamics were inconclusive. In the future, expandingthis collection to include a greater number of strains from a diversearray of ecological niches, including more clinical isolates, as wellas the evaluation of growth kinetics using realistic conditions willhelp us understand the potential risk of this emerging foodbornepathogen.
124 R. Miled-Bennour et al. / International Journal of Food Microbiology 139 (2010) 116–125
Acknowledgements
The authors would like to thank P. Mariani, J.F. Lecrigny, A. Leflècheand I. Desforges for providing some Cronobacter strains or samples,and L. Guillier, H. Bergis, M. Marault, B. Felix and S. Roussel for helpfuladvice. We would also like to thank C. Chastang (Direction Généralede la Concurrence, de la Consommation, et de la Répression desFraudes, Lyon, France), for providing naturally contaminated PIFsamples. The authors are grateful to MQER, CEB units (Afssa LERQAP)and Alfort National Veterinary School (ENVA) for utilization ofBioscreen and PFGE equipment and Steve Brooks and Judy Kwan ofHealth Canada for DNA sequencing services.
References
Anonymous, 2004. Enterobacter sakazakii and other microorganisms in powderedinfant formula: meeting report. : Microbiological risk assessment series 6. WorldHealth Organisation—Food and Agriculture Organisation of the United Nations,Geneva and Rome.
Anonymous, 2005. Enterobacter sakazakii: une bactérie dans le lait pour bébés. Apropos, 8, AFSSA, Maisons Alfort, France.
Anonymous, 2006a. Enterobacter sakazakii and Salmonella in powdered infant formula:meeting report.Microbiological risk assessment series 10.WorldHealthOrganisation—Food and Agriculture Organisation of the United Nations, Geneva and Rome.
Anonymous, 2006b. Milk andmilk products—detection of Enterobacter sakazakii. ISO/TS22964:2006 and IDF/RM 210:2006. International Organization for Standardization,Geneva, Switzerland.
Arseni, A., Malamou-Ladas, E., Koutsia, C., Xanthou, M., Trikka, E., 1987. Outbreak ofcolonization of neonates with Enterobacter sakazakii. Journal of Hospital Infection 9,143–150.
Baldwin, A., Loughlin, M., Caubilla-Barron, J., Kucerova, E., Manning, G., Dowson, C.,Forsythe, S., 2009. Multilocus sequence typing of Cronobacter sakazakii andCronobacter malonaticus reveals stable clonal structures with clinical significancewhich do not correlate with biotypes. BMC Microbiology 9, 223.
Barbosa, W.B., Cabedo, L., Wederquist, H.J., Sofos, J.N., Schmidt, G.R., 1994. Growthvariations among species and strains of Listeria in culture broth. Journal of FoodProtection 57, 765–769.
Bar-Oz, B., Preminger, O., Peleg, A., Block, C., Arad, I., 2001. Enterobacter sakazakii in thenewborn. Acta Pediatrica 90, 356–358.
Bergis, H., Beaufort, A., Cornu, M., Rudelle, S., 2004. Variability of growth of Listeriamonocytogenes in artificially contaminated cold-smoked salmon. 5th ASEPT Interna-tional Conference: Listeria monocytogenes and Risk analysis, March, 17–18th, Laval,France.
Biering, G., Karlsson, S., Clark, N.C., Jónsdóttir, K.E., Lúdvígsson, P., Steingrímsson, O.,1989. Three cases of neonatal meningitis caused by Enterobacter sakazakii inpowdered milk. Journal of Clinical Microbiology 27, 2054–2056.
Block, C., Peleg, O., Minster, N., Bar-Oz, B., Simonh, A., Arad, I., Shapiro, M., 2002. Clusterof neonatal infections in Jerusalem due to unusual biochemical variant ofEnterobacter sakazakii. European Journal of Clinical Microbiology and InfectiousDiseases 21, 613–616.
Breeuwer, P., Lardeau, A., Peterz, M., Joosten, H.M., 2003. Desiccation and heat toleranceof Enterobacter sakazakii. Journal of Applied Microbiology 95, 967–973.
Bruce, J.L., 1996. Automated system rapidly identifies and characterizes microorgan-isms in food. Food Technology 50, 77–81.
Bruce, J.L., Hubner, R.J., Cole, E.M., McDowell, C.I., Webster, J.A., 1995. Sets of EcoRIfragments containing ribosomal RNA sequences are conserved among differentstrains of Listeria monocytogenes. Proceedings of the National Academy of Sciences,USA, 92, pp. 5229–5233.
Clark, N.C., Hill, B.C., O'Hara, C.M., Steingrimsson, O., Cooksey, R.C., 1990. Epidemiologictyping of Enterobacter sakazakii in two neonatal nosocomial outbreaks. DiagnosticMicrobiology and Infectious Disease 13, 467–472.
Cooney, S., Healy, B., O'Brien, S., Fanning, S., Iversen, C., 2009. Growth of Enterobacter-iaceae in milk. First International Conference on Cronobacter (Enterobactersakazakii), January, 22–23th, Dublin, Ireland.
Drudy, D., O'Rourke, M., Murphy, M., Mullane, N.R., O'Mahony, R., Kelly, L., Fischer, M.,Sanjaq, S., Shannon, P., Wall, P., O'Mahony, M., Whyte, P., Fanning, S., 2006.Characterization of a collection of Enterobacter sakazakii isolates from environ-mental and food sources. International Journal of Food Microbiology 110, 127–134.
El-Sharoud, W.M., El-Din, M.Z., Ziada, D.M., Ahmed, S.F., Klena, J.D., 2008. Surveillanceand genotyping of Enterobacter sakazakii suggest its potential transmission frommilk powder into imitation recombined soft cheese. Journal of AppliedMicrobiology105, 559–566.
El-Sharoud, W.M., O'Brien, S., Negredo, C., Iversen, C., Fanning, S., Healy, B., 2009.Characterization of Cronobacter recovered from dried milk and related products.BMC Microbiology 9, 24.
Farmer III, J.J., Asbury, A.A., Hickman, F.W., Brenner, D.J., the Enterobacteriaceae group,1980. Enterobacter sakazakii, new species isolated from clinical specimens.International Journal of Systematic Bacteriology 30, 569–584.
Gadzov, B., Schoder, D., Foissy, H., Malorny, B., Hein, I., Wagner, M., 2006. Molecularcharacterisation of Enterobacter sakazakii to elucidate contamination chains in a milkpowder processing plant. COST 920meeting, September, 11–12th, Ploufragan, France.
Galtier, N., Gouy, M., Gautier, C., 1996. SeaView and Phylo_win, two graphic tools forsequence alignment and molecular phylogeny. Computer Applications in theBiosciences 12, 543–548.
Gnanou Besse, N., Leclercq, A., Maladen, V., Tyburski, C., Lombard, B., 2006. Evaluation ofthe ISO-IDF draft standard method for the detection of Enterobacter sakazakii inpowder infant food formula. Journal of AOAC International 89, 1309–1316.
Guillaume-Gentil, O., Sonnard, V., Kandhai, M.C., Marugg, J.D., Joosten, H., 2005. Asimple and rapid cultural method for detection of Enterobacter sakazakii inenvironmental samples. Journal of Food Protection 68, 64–69.
Gurtler, J.B., Kornacki, J.L., Beuchat, L.R., 2005. Enterobacter sakazakii: a coliformof increasedconcern to infant health. International Journal of Food Microbiology 104, 1–34.
Healy, B., Mullane, N., Collin, V., Mailler, S., Iversen, C., Chatellier, S., Storrs, M., Fanning,S., 2008. Evaluation of an automated repetitive sequence-based PCR system forsubtyping Enterobacter sakazakii. Journal of Food Protection 71, 1372–1378.
Hunter, S.B., Vauterin, P., Lambert-Fair, M.A., Van Duyne, M.S., Kubota, K., Graves, L.,Wrigley, D., Barrett, T., Ribot, E., 2005. Establishment of a universal size standardstrain for use with the PulseNet standardized pulsed-field gel electrophoresisprotocols: converting the national databases to the new size standard. Journal ofClinical Microbiology 43, 1045–1050.
Iversen, C., Forsythe, S., 2003. Risk profile of Enterobacter sakazakii, an emergentpathogen associated with milk formula. Trends in Food Science and Technology 14,443–454.
Iversen, C., Lane, M., Forsythe, S.J., 2004. The growth profile, thermotolerance andbiofilm formation of Enterobacter sakazakii grown in infant formula milk. Letters inApplied Microbiology 38, 378–382.
Iversen, C., Waddington, M., Farmer III, J.J., Forsythe, S.J., 2006. The biochemicaldifferentiation of Enterobacter sakazakii genotypes. BMC Microbiology 6, 94.
Iversen, C., Lehner, A., Mullane, N., Bidas, E., Cleenwerck, I., Marrug, J., Fanning, S., Stephan,R., Joosten, H., 2007. The taxonomy of Enterobacter sakazakii: proposal of a new genusCronobacter gen. nov. anddescriptionsofCronobacter sakazakii comb. nov.Cronobactersakazakii subsp. sakazakii, comb. nov., Cronobacter sakazakii subsp.malonaticus subsp.nov., Cronobacter turicensis sp. nov., Cronobacter muytjensii sp. nov., Cronobacterdublinensis sp. nov. and Cronobacter genomospecies I. BMC Evolutionary Biology 7, 64.
Iversen, C., Mullane, McCardell, B., Tall, B.D.N., Lehner, A.J., Fanning, S., Stephan, R.,Joosten, H., 2008. Cronobacter gen. nov., a new genus to accommodate the biogroupsof Enterobacter sakazakii, and proposal of Cronobacter sakazakii gen. Nov., comb.nov., Cronobacter malonaticus sp. nov., Cronobacter turicensis sp. nov., Cronobactermuytjensii sp. nov., Cronobacter dublinensis sp. nov., Cronobacter genomospecies 1,and of three subspecies, Cronobacter dublinensis subsp. dublinensis subsp. nov.,Cronobacter dublinensis subsp. lausannensis subsp. nov., Cronobacter dublinensissubsp. lactaridi subsp. nov. International Journal of Systematic and EvolutionaryMicrobiology 58, 1442–1447.
Kandhai, M.C., Reij, M.W., Gorris, L.G.M., Guillaume-Gentil, O., van Schothorst, M.,2004a. Occurrence of Enterobacter sakazakii in food production environments andhouseholds. The Lancet 363, 39–40.
Kandhai, M.C., Reij, M.W., van Puyvelde, K., Guillaume-Gentil, O., Beumer, R.R., vanSchothorst, M., 2004b. A new protocol for the detection of Enterobacter sakazakiiapplied to environmental samples. Journal of Food Protection 67, 1267–1270.
Kandhai, M.C., Reij, M.W., Grognou, C., van Schothorst, M., Gorris, L.G., Zwietering, M.H.,2006. Effects of preculturing conditions on lag time and specific growth rate ofEnterobacter sakazakii in reconstituted powdered infant formula. Applied andEnvironmental Microbiology 72, 2721–2729.
Konstantinidis, K.T., Tiedje, J.M., 2007. Prokaryotic taxonomy and phylogeny in thegenomic era: advancements and challenges ahead. Current Opinion in Microbiol-ogy 10, 504–509.
Koort, J.M., Lukinmaa, S., Rantala, M., Unkila, E., Siitonen, A., 2002. Technicalimprovement to prevent DNA degradation of enteric pathogens in pulsed-fieldgel electrophoresis. Journal of Clinical Microbiology 40, 3497–3498.
Kuhnert, P., Korczak, B.M., Stephan, R., Joosten, H., Iversen, C., 2009. Phylogeny andprediction of genetic similarity of Cronobacter and related taxa by multilocussequence analysis (MLSA). International Journal of Food Microbiology 136,152–158.
Lane, D.J., 1985. 16S/23S sequencing. In: Stackbrandt, E., Goodfellow, M. (Eds.), NucleicAcid Techniques in Bacterial Systematics. Wiley, New York, pp. 115–176.
Lebert, I., Bégot, C., Lebert, A., 1998. Development of two Listeria monocytogenes growthmodels in a meat broth and their application to beef meat. Food microbiology 15,499–509.
Lehner, A., Stephan, R., 2004. Microbiological, epidemiological, and food safety aspectsof Enterobacter sakazakii. Journal of Food Protection 67, 2850–2857.
Lenati, R., O'Connor, D., Hébert, K., Farber, J.M., Pagotto, F.J., 2008. Growth and survivalof Enterobacter sakazakii in human breast milk with and without fortifiers, ascompared to powdered infant formula. International Journal of Food Microbiology122, 171–179.
Liesegang, A., Tschäpe, H., 2002. Modified pulsed-field gel electrophoresis method forDNA degradation-sensitive Salmonella enterica and Escherichia coli strains.International Journal of Medical Microbiology 291, 645–648.
Ludwig, W., Strunk, O., Westram, R., Richter, L., Meier, H., Yadhukumar, Buchner A., Lai,T., Steppi, S., Jobb, G., Forster, W., Brettske, I., Gerber, S., Ginhart, A.W., Gross, O.,Grumann, S., Hermann, S., Jost, R., Konig, A., Liss, T., Lussmann, R., May, M., Nonhoff,B., Reichel, B., Strehlow, R., Stamatakis, A., Stuckmann, N., Vilbig, A., Lenke, M.,Ludwig, T., Bode, A., Schleifer, K.H., 2004. ARB: a software environment forsequence data. Nucleic Acids Research 32, 1363–1371.
Membré, J.M., Leporq, B., Vialette, M., Mettler, E., Perrier, L., Thuault, D., Zwietering, M.,2005. Temperature effect on bacterial growth rate: quantitative microbiologyapproach including cardinal values and variability estimates to perform growthsimulations on/in food. International Journal of Food Microbiology 100, 179–186.
125R. Miled-Bennour et al. / International Journal of Food Microbiology 139 (2010) 116–125
Mullane, N.R., Whyte, P., Wall, P.G., Quinn, T., Fanning, S., 2007. Application of pulsed-field gel electrophoresis to characterise and trace the prevalence of Enterobactersakazakii in an infant formula processing facility. International Journal of FoodMicrobiology 116, 73–81.
Mullane, N., Healy, B., Meade, J., Whyte, P., Wall, P.G., Fanning, S., 2008a. Disseminationof Cronobacter spp. (Enterobacter sakazakii) in a powdered milk proteinmanufacturing facility. Applied and Environmental Microbiology 74, 5913–5917.
Mullane, N.R., Ryan,M., Iversen, C., Murphy,M., O'Gaora, P., Quinn, T.,Whyte, P.,Wall, P.G.,Fanning, S., 2008b. Development of multiple-locus variable-number tandem-repeatanalysis for the molecular subtyping of Enterobacter sakazakii. Applied andEnvironmental Microbiology 74, 1223–1231.
Muytjens, H.L., Roelofs-Willemse, H., Jaspar, H.J., 1988. Quality of powdered substitutesfor breast milk with regard to members of the family Enterobacteriaceae. Journal ofClinical Microbiology 26, 743–746.
Nazarowec-White, M., Farber, J.M., 1997a. Enterobacter sakazakii: a review. Interna-tional Journal of Food Microbiology 34, 103–113.
Nazarowec-White, M., Farber, J.M., 1997b. Incidence, survival and growth ofEnterobacter sakazakii in infant formula. Journal of Food Protection 60, 226–230.
Nazarowec-White, M., Farber, J.M., 1999. Phenotypic and genotypic typing of food andclinical isolates ofEnterobacter sakazakii. Journal ofMedicalMicrobiology48, 559–567.
Proudy, I., Bouglé, D., Leclercq, R., Vergnaud, M., 2008a. Tracing of Enterobacter sakazakiiisolates in infant milk formula processing by BOX-PCR genotyping. Journal ofApplied Microbiology 105, 550–558.
Proudy, I., Bouglé, D., Coton, E., Coton, M., Leclercq, R., Vergnaud, M., 2008b. Genotypiccharacterization of Enterobacter sakazakii isolates by PFGE, BOX-PCR andsequencing of the fliC gene. Journal of Applied Microbiology 104, 26–34.
Ribot, E.M., Fair,M.A., Gautom,R., Cameron,D.N., Hunter, S.B., Swaminathan, B., Barrett, T.J.,2006. Standardization of pulsed-field gel electrophoresis protocols for the subtypingof Escherichia coli O157:H7, Salmonella, and Shigella for PulseNet. FoodbornePathogens and Disease 3, 59–67.
Rosset, P., Noel, V., Morelli, E., 2007. Time-temperature profiles of infant milk formula inhospitals and analysis of Enterobacter sakazakii growth. Food Control 18, 1412–1418.
Silbert, S., Boyken, L., Hollis, R.J., Pfaller, M.A., 2003. Improving typeability of multiplebacterial species using pulsed-field gel electrophoresis and thiourea. DiagnosticMicrobiology and Infectious Disease 47, 619–621.
Simmons, B.P., Gelfans, M.S., Haas, M., Metts, L., Ferguson, J., 1989. Enterobacter sakazakiiinfections in neonates associated with intrinsic contamination of a powdered infantformula. Infection Control and Hospital Epidemiology 10, 398–401.
Van Acker, J., De Smet, F., Muyldermans, G., Bougatef, A., Naessens, A., Lauwers, S., 2001.Outbreak of necrotizing enterocolitis associated with Enterobacter sakazakii inpowdered milk formula. Journal of Clinical Microbiology 39, 293–297.
