Molecular characterisation of Salmonella strains by an oligonucleotide multiprobe microarray

ARTICLE IN PRESS

0890-8508/$ - se

doi:10.1016/j.m

�CorrespondE-mail addr

Molecular and Cellular Probes 21 (2007) 56–65

www.elsevier.com/locate/ymcpr

Molecular characterisation of Salmonella strains byan oligonucleotide multiprobe microarray

Burkhard Malornya,�, Cornelia Bungea, Beatriz Guerraa, Sandra Prietzb, Reiner Helmutha

aNational Salmonella Reference Laboratory, Federal Institute for Risk Assessment, Diedersdorfer Weg 1, D-12277 Berlin, GermanybScienion AG, VolmerstraX e 7b, D-12489 Berlin, Germany

Received 12 January 2006; accepted 3 August 2006

Available online 17 September 2006

Abstract

A DNA microarray has been developed for the simultaneous characterisation and typing of Salmonella enterica subsp. enterica

isolates. One-hundred and nine 35–40mer oligonucleotides probes detect flagellar and somatic antigen encoding genes (serogroup or

serotype specific), important virulence genes located within or outside the pathogenicity islands, phage-associated genes and antibiotic

resistance determinants. The probes were printed on glass slides and whole genomic Cy5-labelled Salmonella DNA was hybridised to the

substrate. A set of 19 different Salmonella strains and one Escherichia coli strain has been selected as positive and negative controls for

each probe. The validity of the results is confirmed by gene-specific PCRs or phenotypic methods (serotyping, MIC determination for

various antimicrobial agents). Of 2071 data points generated, an agreement of 97.4% has been obtained between microarray and PCR/

phenotypic results. Twenty-six data points (1.3%) were classified as uncertain and, similarly, 1.3% showed a discordant result. The

microarray described here is a new tool to study the epidemiology of Salmonella strains on the genotypic level and might become a

powerful method in risk assessment studies.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Microarray; Salmonella; 40mer oligonucleotides; Characterisation

1. Introduction

Taxonomically the genus Salmonella is divided into twospecies, Salmonella enterica and Salmonella bongori, eachof which contains multiple serovars [1]. S. enterica

comprises seven subspecies. Of these special subspeciesenterica serovars are associated with humans and warm-blooded animal infections. The differentiation of Salmo-

nella in serotypes is based on their antigenic variation in thelipopolysaccharide (O-antigen) and flagellae (H1- and H2-antigens). Serovars can differ in their pathogenicity andhost range. In addition they show considerable variabilityin resistance to a broad spectrum of antibiotics.

With the availability of complete Salmonella genomesequences, whole genome analyses become possible byidentifying the gain, loss and divergence of genes betweendifferent lineages of Salmonella [2]. Comparative genomic

e front matter r 2006 Elsevier Ltd. All rights reserved.

cp.2006.08.005

ing author. .: +49 30 8412 2237; fax: +4930 8412 2953.

ess: [email protected] (B. Malorny).

hybridisations using microarray technology between the S.

enterica ssp. enterica serotype Typhimurium LT2 genomeand other subspecies strains of Salmonella have beensuccessfully applied using PCR products as probes,revealing differences in hundreds of genes [2–5]. Anonredundant microarray of S. enterica serovar Typhi-murium LT2 and Typhi CT18 has been applied to assessthe genomic content of diverse isolates of serovar Typhi [6].Despite the high clonality of Typhi, it was shown that thegenomic reservoir is unstable [6]. This indicates that lateralgene transfer is a major contributor to Salmonella

evolution [7]. The various Salmonella genomes containhorizontally acquired genetic elements that might play arole in infection, host adaptation and disease development.The different serogroups (O-antigens) of Salmonella are

primarily based on the different gene content of the rfb

region encoding different core and sidechain lipopolysac-charide. The different types of the structural unit offlagellae (H1- and H2-antigens) are encoded by the fliC

(H1) and fljB (H2) genes. They harbour variable regions

www.elsevier.com/locate/ymcpr

dx.doi.org/10.1016/j.mcp.2006.08.005

mailto:[email protected]

ARTICLE IN PRESSB. Malorny et al. / Molecular and Cellular Probes 21 (2007) 56–65 57

which are responsible for the various types of flagellaeantigens.

Plasmids are extrachromosomal, self-transmissible andself-circular DNA molecules that vary in size between twokb and several hundred kb. They can encode virulencegenes such as the spv locus or fimbrial subunits (pef) [8],and also genes conferring antimicrobial resistance. Phagesare mobile elements which are mediated by transduction.In Salmonella many phages are integrated as prophages inthe Salmonella genome. It was shown that they can encodeautonomously expressed virulence genes [9]. A prominentphage is SopEF encoding the effector protein SopE [10].Other mobile genetic elements are the Salmonella patho-genicity islands (SPIs). Five SPIs had been identified in S.Typhimurium [11]. They are characterised by their absencefrom the Escherichia coli genome, the G+C content andthe impact of genes on the pathogenicity of Salmonella.Briefly, SPI1 and SPI2 encode two different type IIIsecretion systems and several effector proteins such asavrA, sptP, slrP, sipA, sseF and sseG. SPI2 confers survivalin macrophages, SPI3 encodes a magnesium transportsystem necessary for full virulence and macrophagesurvival. SPI4 encodes a putative type 1 secretionsystem and SPI5 encodes the effector protein sopB [11].Salmonella contain many different fimbrial gene clusters.They encode surface hair-like structures necessary for theprimary attachment to the host cells and survival inparticular ecological niches. Different serotypes canharbour different subsets of these clusters [4]. The lack orpresence of certain fimbrial clusters might be a sign todifferent environmental adaptations or selection pressure.For S. Typhimurium LT2 11 different fimbrial clusters areknown [12].

During the last decade the understanding of themolecular basis and detection methods for the spread ofantimicrobial resistance in Salmonella has developedtremendously [13]. Resistance genes can move betweenchromosomal and extra-chromosomal DNA elements andthey may move between bacteria of the same or differentspecies or to bacteria of different genera by horizontal genetransfer. The most important vehicles for transfer ofresistance genes in bacteria are mobile genetic elements,such as plasmids, transposons, integrons, gene cassettesand genomic islands [14].

However, the use of a DNA microarray reflecting thegenome of one isolate limits the comprehensive character-isation of related genomes. While the absence of genes canbe detected in test strains, genes that are unique for the teststrain cannot be monitored. In addition, the use of PCRproducts as hybridisation probes has the disadvantage oflow hybridisation specificity due to high nucleotidehomology between gene family members. Highly variableregions within a gene family could not be discriminated byPCR product-based microarrays. To increase hybridisationspecificity, the use of smaller length probes (i.e. synthesised50mer oligonucleotides) is meanwhile becoming a con-venient method for spotting [15]. In addition, the produc-

tion of oligo-based microarrays requires fewer resourcesthan PCR product-based microarrays.The aim of this study was the development of a DNA

microarray useful for the molecular characterisation andtyping of Salmonella isolates by combining variousimportant genetic markers selected from various Salmo-

nella strains. The DNA microarray contains oligonucleo-tide probes for detecting genes encoding antibioticresistance, pathogenicity, fimbriae, phage-associated genes,flagellae (H-antigens), lipopolysacharides (O-antigens) andsome others.

2. Materials and methods

2.1. Selection of the target genes and probe design

The aim was to design oligonucleotide probes whichidentify target sequences within a Salmonella genomegiving information on genetic determinants within aparticular isolate. This was based on the detection of theserotype, fimbrial clusters, important virulence genes,phage-associated genes and genes encoding antimicrobialresistance or complex resistance determinants. We havelimited the number of targets (probes) firstly, to evaluatethe usefulness of the microarray approach, and secondlynot to provide the end-user with uninformative data.Relevant open reading frame sequences were selected

from the Nucleotide NCBI Genebank (http://www.ncbi.nlm.nih.gov) and imported in Array Designer 2.0(Premier Biosoft, Palo Alto, CA) followed by a crosshomology analysis against the genome sequence of strainS. enterica ssp. enterica serotype Typhimurium strain LT2(Accession no. NC_003197). Based on the avoidance ofcross homologies, 109 target gene-specific 35–40meroligonucleotide probes were designed by the programusing the recommended default options for 40mer oligo-nucleotides. Many of the probes indicate the specificpresence or absence of a Salmonella encoded gene, otherprobes were located within variable regions of one gene(such as fliC and fljB). The probe sequences, accessionnumbers and its functional characteristics are shown inTable 2. On request additional oligonucleotide probe dataare available.

2.2. Selection of bacterial reference strains

A set of 19 S. enterica subsp. enterica and one E. coli

strain were selected from the strain collection of theNational Salmonella Reference Laboratory Berlin, Ger-many (Table 1). These strains represented positive andnegative controls for all oligonucleotide probes printed onthe microarray except the tet(E) and aacC1 genes for whichno reference strain was available. The E. coli strain wasselected as a positive control for the tet(D) gene. S.

Typhimurium strain LT2 was defined as the referencestrain since its genome has been sequenced completely [12].

http://www.ncbi.nlm.nih.gov

http://www.ncbi.nlm.nih.gov

ARTICLE IN PRESS

Table 1

Reference strains used in this study for microarray hybridisations

Reference strain no. Serotype Phage type Serogroup Antigenic

formula

Phenotypic antimicrobial resistance

profileb

LT2 Typhimurium B 1,4,12:i:1,2 Susceptible

SUO5 Typhimurium DT120 B 1,4,12:i:1,2 AMP CHL SPE STR SUL TET

SUO1 Typhimurium DT104 B 1,4,12:i:1,2 AMP CHL FLO SPE STR SUL TET

NRL 00-419 Typhimurium DT104 B 1,4,12:i:1,2 AMP CHL FLO NAL SPE STR SUL

TET TMP SXT

SUO8 [4,5,12:i:-] U302 B 4,5,12:i:- AMP CHL GEN SPE STR SUL TET

TMP SXT

SUO6 Typhimurium RDNCa B 1,4,12:i:1,2 AMP CHL KAN NEO SPE STR SUL

TET TMP SXT

S65 (NRL 01-02571) Typhimurium DT104A B 1,4,12:i:1,2 AMP CHL SPE STR SUL TET TMP

SXT

S40 (NRL 01-01338) Typhimurium DT12 B 1,4,12:i:1,2 TET

NRL 99-4068 Typhimurium DT120 B 1,4,12:i:1,2 AMP CHL GEN KAN NEO SPE STR

SUL TET TMP SXT

NRL 98-3363 Schleissheim B 4,12,27:b:- Susceptible

NRL 01-1543 Paratyphi B (d-

tartrate+)

B 1,4,12:b:1,2 AMP AMC NAL SPE STR SUL TMP

SXT

NRL 01-1380 Saintpaul B 1,4,12:e,h:1,2 AMP GEN KAN NAL NEO SPE STR

SUL TET

NRL 99-601 Hadar C2-C3 6,8:z10:e,n,x AMP NAL STR TET

NRL 01-2132 Goldcoast C2-C3 6,8:r:l,w GEN STR SPE SUL TET

NRL 00-4 Enteritidis D1 1,9,12:g,m:- Susceptible

RKI-Ty1 Typhi London D1 9,12,Vi:d:- Susceptible

NRL 03-1949 Lindern H 6,14,24:d:e,n,x Susceptible

NRL 02-102 Oranienburg C1 6,7:m,t:- Susceptible

NRL 99-929 Anatum E1 3,10:e,h:1,6 Susceptible

NRL EC227 E. coli AMP CHL KAN NEO STR SUL TET

TMP SXT

NRL: National Reference Laboratory for Salmonella; SUO: Salmonella University of Oviedo; RKI: Robert Koch Institute.aRDNC reaction does not conform.bAMP, ampicillin; AMC, amoxicillin/clavulanic acid; CHL, chloramphenicol; FLO, florfenicol; GEN, gentamicin, KAN, kanamycin; NEO, neomycin;

NAL, nalidixic acid; SPE, spectinomycin; STR, streptomycin; SMX, sulphamethoxazole; SXT, trimethoprim/sulphamethoxazole; TET, tetracycline;

TMP, trimethoprim.

B. Malorny et al. / Molecular and Cellular Probes 21 (2007) 56–6558

2.3. DNA purification

Salmonella strains were grown for 16–18 h at 37 1C inLuria-Bertani medium with gentle shaking. A 1ml culturealiquot was taken for genomic DNA isolation usingDNeasy Tissue Kit (Qiagen, Hilden, Germany) accordingto the manufacture’s protocol. The amount of DNA wasspectrophotometrically determined by measuring the op-tical density at 260 nm.

2.4. Fluorescence labelling of genomic DNA

The genomic DNA was labelled with Cy5 by usingDecaLabel DNA labelling kit (MBI Fermentas, Vilnius,Lithuania). Reaction mixtures of 45 ml consisted of 4 mgDNA and 10 ml of decanucleotides in reaction buffer weredenatured at 95 1C for 5min and immediately placed onice. A 3 ml aliquot of deoxynucleotide Mix C (containing nodCTP), 1 ml of Cy5-dCTP (Amersham Biosciences, Frei-burg, Germany) and 1 ml of the Klenow fragment (5U/ml)was added and the labelling reaction mixture incubated at37 1C for 30min. The addition of 4 ml dNTP-Mix and an

incubation at 37 1C for 15min followed by inactivation ofthe Klenow fragment at 65 1C for 15min finalised thelabelling reaction. Unincorporated deoxynucleotides anddecanucleotides were removed by using the Mini ElutePCR Purification kit (Qiagen). An additional step beforeapplying Elution buffer on the column was performed byadding 400 ml of a 35% (w/v) guanidinium chloridesolution on the column followed by a 1min centrifugationstep at high speed. Finally, the eluted DNA was vacuumdried and stored on ice in the dark until use.

2.5. Microarray manufacture

Oligonucleotide probe synthesis and microarray printingwere performed at Scienion AG (Berlin, Germany). The 50

terminal nucleotide of each oligonucleotide was aminatedto allow a covalently coupling to the aldehyde groupscoated on glass slides (Scienion AG). Oligonucleotideswere printed in a concentration of 50 mM using SciSpot-Oligo printing buffer (Scienion AG) on SuperAldehydesubstrates (TeleChem International Inc., Sunnyvale, CA)in duplicate. On one slide two arrays were printed. Each


array consists of two subarrays. One subarray consists ofeight blocks (six columns and four rows). The diameter of aspot was approx. 200 mm. A set of nine different 70meroligonucleotides derived from the Arabidopsis thaliana

genes Cab (X56062), RCA (X14212), rbcL (U911966),LPT4 (AF159801), LPT6 (AF159803), XCP2 (AF191028),RCP1 (AF168390), NAC1 (AF19805) and PRKASE(X58149) (Stratagene La Jolla, CA) were chosen asnegative controls. Per subarray, each probe was printedin duplicate except for gene LPT6. LPT6 and spottingbuffer were printed on the last row of each of the eightblocks. The customised ready-to-use Salmonella arrayswere stored in the dark at 8 1C and used within 3 months.

2.6. Hybridisation and posthybridisation washing

Vaccuum dried Cy5 labelled genomic DNA was resus-pended in 30 ml prewarmed sciHYB Hybridisation buffer(Scienion AG) containing 50% (v/v) formamide anddenatured at 98 1C for 2min. After cooling to roomtemperature in the dark the complete solution was carefullypipetted onto the Salmonella array, on which a supportedcoverslip (Erie Scientific, Portsmouth, NH) had beenplaced before. The slide was inserted into a hybridisationchamber (Scienion AG) containing four times 20 ml double-distilled water for subsequent humidity equilibration.Hybridisation was performed overnight (18–20 h) at 42 1Cin a water bath. After hybridisation, the slides were washedby gentle shaking immediately under stringent conditionsfor 3min at room temperature in 200ml washing buffer 1(containing 1 X SSC, and 0.3% SDS). After a second washfor 2min with washing buffer 2 (0.2 X SSC), a finalwashing step followed by using washing buffer 3 (0.05 XSSC). The slide was dried by gently centrifugation for 3minat 300 g and stored until scanning in the dark at roomtemperature. All reference strain DNAs (Table 1) exceptstrain 01-1380 were independently hybridised on twoarrays generating four spot intensities for each probe.

2.7. Scanning and analysis

The hybridised slides were scanned with a GenePix4000B laser scanner (Axon, Foster City, CA) at aPhotomultiplier tube (PMT) gain of 800–900 using laserlight of wavelength at 635 nm to excite the Cy5 dye.Fluorescent images were captured as multi-image-taggedimage file format and analysed with GenePix Pro 4.1Software (Axon). For the normalisation of the Cy5 probesignals a ratio has been calculated. The local backgroundsubtracted median spot intensity of each probe was dividedby the local background subtracted median spot intensityof the Salmonella probe for the ttrC gene, which is presentin all S. enterica subspecies strains [16]. Based on the spotintensities of all negative target probes of S. Typhimuriumstrain LT2 probe signals, for which the ratio was equal toor greater than 0.4, were considered as positive. Ratiovalues between 0.3 and 0.4 were classified as ‘‘uncertain’’.

For the E. coli reference strain EC227, an artificial value of10.000 units fluorescence for the ttrC probe was applied fornormalisation since the ttrC probe gave no signal.

2.8. Validation of microarray signals with PCR

PCRs were performed for the target genes listed in Table2. PCR primers resulting in 400–500 bp products weredesigned by the Array Designer 2.0 (Premier Biosoft, PaloAlto, CA). For the detection of the most antibioticresistance genes, published primer sequences from varioussources were used. A table of PCR primers, PCR productsizes and references can be obtained on request. A typicalPCR (25 ml) contained 0.4 mmol l�1 of each primer,200 mmol l�1 of each dNTP, 1 X PCR reaction buffer(20mmol l�1 Tris-HCl (pH 8.4), 50mmol l�1 KCl),1.5mmol l�1 MgCl2, 1U Platinum Taq polymerase (Invi-trogen, Karlsruhe, Germany) and 1 ng of the same purifiedDNA, as used for the Cy5-labelling. The incubationconditions were 95 1C for 1min, followed by 33 cycles of95 1C for 30 s, 55–60 1C for 30 s and 72 1C for 30 s. A 10 mlaliquot of a PCR product was loaded on a 1.5% agarosegel and electrophoresed at 6V cm�1 for 90min. Thepresence of a clear fragment with the correct size afterstaining the gel in ethidiumbromid has been assessed aspositive signal (presence of the gene).

2.9. Supplemental data

Supplemental data of the results of DNA microarrayhybridisations and a list with the PCR primer sequencesused for the identification of specific genes in Salmonella

strains can be downloaded from the MCP homepage.

3. Results and discussion

3.1. Construction of the microarray

One hundred and nine 35–40mer-oligonucleotide probeswere designed for the molecular characterisation andtyping of Salmonella isolates (Table 2). Although50–70mer oligonucleotides [17] or PCR products [18] arepreferably described for microarray-based expressionstudies, in this study we focus on 35–40mer oligonucleo-tides for comparative genomic hybridisations in order toensure maximum specificity for the discrimination of shortvariable regions between different alleles. Each oligonu-cleotide represents the presence/absence of a characteristicsequence previously specifically or commonly identified invarious Salmonella serotypes. We have focused on themarker groups serotypes, fimbrial clusters, pathogenicity,phage-associated genes, resistance and others (mainlymetabolic pathway-associated genes). The number oftarget sequences has been limited, firstly, to evaluate theusefulness of the microarray approach on a defined set oftargets, and secondly not to provide the end-user withuninformative data. Previously, only thematic Salmonella

ARTICLE IN PRESSB. Malorny et al. / Molecular and Cellular Probes 21 (2007) 56–6560

DNA microarrays (mainly antibiotic resistance microar-rays) using PCR products [19,20] or oligonucleotides [21]as DNA probes were described in the literature.

3.2. Validation and normalisation of the microarray signals

Random primed labelled genomic DNA of 19 Salmo-

nella reference strains and one E. coli strain (Table 1) wereapplied to the microarray. Since it was previously shownthat the ttrC gene is present in all S. enterica subspecies[16], the ttrC specific gene probe was selected for thenormalisation of the Cy5-microarray signals. This allowedto calculate a ratio of each specific probe signal with thettrC probe signal based on the presence of the targetsequences within the S. Typhimurium LT2 genome [12]. Aratio of p0.3 indicated as the absence, and a ratio of X0.4as the presence of a probe sequence. Values between 0.3and 0.4 were classified as uncertain.

The validity of the threshold levels was confirmed withthe results of a comprehensive PCR screening. For eachtarget sequence, except those from the serotype markergroup, primers were constructed or selected from literatureand used for PCR screening to identify the presence orabsence of the target sequences among the 19 Salmonella

reference and one E. coli strain (Table 1). The list of primersequences can be obtained from the MCP homepage.Probes of the serotype-specific marker group were com-pared with the results of the classical serotyping accordingto the Kauffmann–White scheme.

Negative controls derived from nine A. thaliana genesand spotting buffer controls were all below a ratio of 0.1(data not shown) and classified as negative. Supplementaldata are available on request showing the ratios of eachspecific probe signal with the ttrC probe signal of thereference strains tested.

3.3. Microarray analysis

Fig. 1 summarises the genotypic characteristics of the 20reference strains tested, determined by microarray, PCRand traditional serotype testing. Of the 2071 data pointsobtained from the 19 Salmonella reference strains, 1.3%(26 data points) were classified as ‘‘uncertain’’ and 1.3%(27 data points) gave a discordant result between themicroarray and the PCR/serotyping results. The discre-pancies are preferably caused by two probes, the abeC2-C3

(five false positive microarray signals) and the spvC probe(six false positive microarray signals). An additional probe(stdB) showed four ambiguous (uncertain) and onediscordant result (PCR positive, microarray signal nega-tive). Obviously, these probes generated an unspecific or tolow signal, and these probes need redesigning. However,not all discrepancies are necessarily caused by a false signalof the microarray. For example, the S. Typhi strain RKI-Ty1 tested expresses the Vi antigen. Surprisingly, animportant gene in the Vi polysaccharide synthesis, thewzf gene, could not be detected by PCR but with the DNA

microarray. This indicates a false negative result of thePCR probably caused by polymorphic sites within the PCRprimers.

3.3.1. Serotype specific genes (flagellar and somatic genes)

Nine probes discriminate variable regions of the fliC andfljB genes encoding the structural proteins of the H1- andH2-antigens. They are expressed by serotypes with highepidemiological importance worldwide. Due to highlyhomologues sequences between the specific alleles theprobe sequences identify rather related H antigens(Table 2). Consequently, highly homologue fliC or fljB

alleles encoding similar antigens (e.g.1-complex including1,2; 1,5; 1,6; 1,7) [22] cannot be discriminated by the35–40mer probes. One probe whose sequence is locatedwithin a 50-conserved region of the fliC gene identified allfliC and fljB alleles correctly. Reference strains SUO8 (S.[4,5,12:i:-]), 98-3363 (S. Schleissheim), 00-4 (S. Enteritidis),RKI-Ty1 (S. Typhi) and 02-102 (S. Oranienburg) do notexpress the second-phase antigen encoded by the fljB gene.For all of these reference strains, the microarray datarevealed the absence of the hin gene, which encodes a DNAinvertase responsible for the regulation of flagellar geneexpression. For the fliC_d probe, the signal ratio for two S.Typhimurium strains (LT2 and SUO1) was slightly overthe cut of level of 0.4 indicating a false positive signal.However, the fljB_1,x and fliC_i probe (specific for all S.Typhimurium strains) gave the correct positive signals. Thereason for this cross-signal might be inaccuracies duringthe hybridisation or printing process. Significant sequencehomologies between the probe sequences do not exist.Genes involved in specific LPS synthesis of Salmonella

serogroups were all identified correctly as describedpreviously and confirmed by published DNA sequences[23]. For example, the wbaV (D1) probe gave only with theserogroup D1 strains a signal (serotype Typhi andEnteritidis). The wbaO (E1, D2) probe specific forserogroups E1 and D2 gave no signal with these serotypes.The probe abe (B) detected specifically all serogroup Bisolates. Two probes wzx (O6,14) and wzy a(1_6) identifiedspecifically the O-antigen factors O27 (S. Schleissheim 98-3363) and O6,14 (S. Lindern 03-1949), respectively.

3.3.2. Fimbrial clusters

Salmonella can possess many different fimbrial geneclusters [7]. Ten probes were constructed which identifyDNA of 10 different fimbrial clusters present in S.

Typhimurium strain LT2. All probes gave a signal withall other S. Typhimurium strains as well as with the S.

Hadar and S. Saintpaul reference strain tested. For theserotype Enteritidis strain (00-4) a similar number of genesencoding fimbriae could be identified. Instead of the stj

cluster, it contains the sef fimbrial cluster. Other serotypeslacked up to eight of these signals (S. Schleissheim 98-3363). These data indicate that different serotypes harbourdifferent subsets of fimbrial clusters. The lack or presenceof certain fimbrial clusters might be functionally relevant

ARTICLE IN PRESSTable

2

Oligonucleotideprobes

forthecharacterisationandtypingof

Sa

lmo

nel

laisolates

Accessionnumber

Probename

Marker

groupGenefunction

Controlstrain

Probesequence

Length

Meltingtemp(1C)

AJ009820

aadA1a

Resistance

Aminoglycoside-300-adenyltransferase:STR/SPEresistance

SUO5

GTGGTGATCGCCGAAGTATCGACTCAACTATCAGAGGTAG

40

69.9

AF261825

aadA2

Resistance


SUO1

ATTTCGAACCAACTATCAGAGGTGCTAAGCGTCATTGAGC

40

69.8

AF169041

aadA5

Resistance


00-419

CGTTCGATATGCCAAAGCAACGATTGAGAGAATCTTGCGT

40

70.5

AF078527

aadB

Resistance

Aminoglycoside-200-adenyltransferase:GEN

resistance

99-2175

CTTGGTGGGCAGACGAAGCGTATGAAATCGCGGAG

35

72.1

X01385

aac(3)-IV

Resistance

Aminoglycoside-3-acetyltransferase:GEN

resistance

SUO8

CCGTTACACCGGACCTTGGAGTTGTCTCTGACACATTCTG

40

71.9

AJ310480

aacC

1Resistance


resistance

Notavailable

CAGTGGCTCTCTATACAAAGTTGGGCATACGGGAAGAAGT

40

70.2

S68058

aacC

2Resistance


resistance

99-4068

GGCATCAGAATACGATTCAAACGGCATTCTCGATTGCTTT

40

69.5

X61367

tet(A)

Resistance

Effluxpump:TETresistance

SUO8

CCTCCTCTTCACGGCGATCTATGCGGCTTCTATAACAACG

40

71.9

V00611

tet(B)

Resistance


SUO5

CTTTGGATGCTGTATTTAGGCCGTTTGCTTTCAGGGATCA

40

70.2

J01749

tet(C)

Resistance


S40

CCTACTACTGGGCTGCTTCCTAATGCAGGAGTCGCATAAG

40

71.5

L06798

tet(D)

Resistance


EC227

AAGCGCAGGTATCAGCTTTATCACACTGCTTAAACCTCTG

40

69.8

L06940

tet(E)

Resistance


Notavailable

AGTGTGGGTGTTGTATTTGGGACGCTTAATTGCTGGTATT

40

69.9

S52437

tet(G)

Resistance


SUO1

CTCTATCGCAGGACCGCTTGGCTTCACAGCACTCTATTCT

40

72.7

AY103456

dfrA1like

Resistance

Dihydrofolate

reductase:TMPresistance

01-1543

TGCGGTCGTAACACGTTCAAGTTTTACATCTGACAATGAG

40

69.0

AF175203

dfrA12

Resistance

Dihydrofolate


SUO8

ACCATTCAAGCTGTAATTCCGTACACCCACTCCGTTTATG

40

69.7

AF220757

dfrA17-7

Resistance

Dihydrofolate


00-419

TGGGTGTTCTTCCAAATCGCAAATATGCAGTAGTGTCAAA

40

68.9

AJ313522

dfrA14

Resistance

Dihydrofolate


SUO6

ACCCGCTCAGGTTGGACATCAAATGATGACAATGTAGTTG

40

70.0

M28829-2

strB

Resistance

Aminoglycoside-600-phosphotransferase:STR

resistance

S65

ATTGAAACCTATAGAAGACATTGCTGATGAACTGCGCGGG

40

70.1

M28829-1

strA

Resistance

Aminoglycoside-300-phosphotransferase:STR

resistance

S65

CATGCCGAACTTCATGGTGGACCCTAAAACTCTTCAATGC

40

70.6

M64556

cm1A1

Resistance

Effluspump:CHL-resistance

SUO8

ACCGAAATATGGGCTTTGCAGTCCGTGTTAGGCTTTATTG

40

70.3

AF118107

floR

Resistance

Putativeeffluxprotein:FLO/C

HLresistance

SUO1

CCGCTATGATCCAACTCACGTTGAGCCTCTATATGGTGAT

40

70.1

X12870

int1

Resistance

DNA

Integrase1:integronassociated

SUO1

GCATTACAGCTTACGAACCGAACAGGCTTATGTCCACTGG

40

71.3

L10818

int2

Resistance

DNA

Integrase2:integronassociated

01-1543

AGGGCATAACGATGTTAAGACCACGCAAATCTATACGCAT

40

69.5

X12869

sul1

Resistance

Dihydropteroate

synthase:SULresistance

SUO1

GATCGAAATGCTGCGAGTCGGATCAGACGTCGTGGATG

38

72.5

M36657

sul2

Resistance

Dihydropteroate


SUO5

CTCAAGGCAGATGGCATTCCCGTCTCGCTCGACAGTTATC

40

73.2

AJ459418

sul3

Resistance

Dihydropteroate


S65

ACTGGAACCGATGTGAAATCTCGTTTAGCACCAACTCTTG

40

69.9

AF261825

qacE

D1

Resistance

Qacmultidrugexporter:Et-Brandquaternary

ammonium

resistance

SUO1

TATCGCAATAGTTGGCGAAGTAATCGCAACATCCGCATTA

40

69.9

K03089

merA

Resistance

Hg(II)

reductase:Mercury

resistance

SUO5

CCTGCGTCAATGTCGGTTGTGTGCCGTCCAAGATCATG

38

73.3

AJ238349

bla

oxa�1like

Resistance

Beta-lactamase:AMPresistance

SUO5

AGAAACAACGGATTAACAGAAGCATGGCTCGAAAGTAGCT

40

69.7

AJ238349

bla

oxa�1like

Resistance


SUO5

ACTGTCGCATCTCCATTATTTGAAGGAACTGAAGGTTGTT

40

68.4

AF153200

bla

pse�1

Resistance


SUO1

CTCGGAGTATTACAGCAGTTGTGTGGAGTGAGCATCAAGC

40

71.0

AF153200

bla

pse�1

Resistance


SUO1

AAGTTTCTCTTTCTGCTCGTATAGGTGTTTCCGTTCTTGA

40

67.9

AF309824

bla

tem�1like

Resistance


SUO8

AATTAATAGACTGGATGGAGGCGGATAAAGTTGCAGGACC

40

69.3

AF309824

bla

tem�1like

Resistance


SUO8

GAGTGATAACACTGCGGCCAACTTACTTCTGACAACGATC

40

70.0

AF024666

aphA1

Resistance

Aminoglycoside-300-phosphotransferase:KAN

resistance

99-4068

AGGTAGCGTTGCCAATGATGTTACAGATGAGATGGTCAGA

40

70.1

M90846

invA

Pathogenicity

SPI1encoded

invasionprotein

LT2

GTGCTGCTTTCTCTACTTAACAGTGCTCGTTTACGACCTG

40

70.0

AE008831

orgA

Pathogenicity

SPI1encoded

oxygen-regulatedinvasionprotein

LT2

TTGTTGAAACGGCAGTAGGCGTCATTAAGCATCATCTTGA

40

70.0

AE008831

hilA

Pathogenicity

SPI1encoded

transcriptionactivator

LT2

AAGAACATGCGATTAAGGCGACAGAGCTGGACCACAATAA

40

70.8

AF013573

avrA

Pathogenicity

SPI1encoded

protein

LT2

CGAAGCATTGACCTGTATTGTTGAGCGTCTGGAAAGTGAA

40

70.2

AF282268

ttrC

Pathogenicity

SPI2encoded:Tetrathionate

reductase

subunitC

LT2

ATGACGCATTCACTCATCATTGAAGAAGTGCTGGCTCACC

40

71.4

AE008761

ssaQ

Pathogenicity

SPI2encoded

protein:Secretionsystem

apparatus

LT2

GGAACGTCTTCAGTCGAACTTGAGCAGATACCACAACAGG

40

70.9

AE008761

ssrB

Pathogenicity

SPI2encoded

protein:Secretionsystem

regulator

LT2

GCCTGTTGTGCATACGAGCCTGACATACTTATCCTTGATC

40

69.8

AF106566

mgtC

Pathogenicity

SPI3intramacrophagesurvivalandgrowth

inlow

Mg2+

LT2

ACGATAATATCACCGCAATTCACTGGAGCATTGATAGTCA

40

68.0

AE008875

misL

Pathogenicity

SPI3encoded

protein

LT2

CGCAGCAGAACAGTACGATTAAGCTCAATATGACGGACAA

40

69.8

AF106566

marT

Pathogenicity

SPI3encoded

putativetranscriptionalregulator

LT2

CGGCATGGTGGTGTCGGCAAATACGCTTTATCAGAATATC

40

70.5

AF060869

spi4_D

Pathogenicity

SPI4encoded

protein

LT2

AAGGAGGGACGATACAAGATATTTATGTAGCCGAGGGTGA

40

69.2

AF060869

spi4_R

Pathogenicity

SPI4encoded

protein

ofABC

transporter

family

LT2

GGGCGTGAGTTATCAGTATGATGCTCAATCTCCGATGATT

40

69.3

AE008747

pipA

Pathogenicity

SPI5encoded

protein

LT2

AAACTACCGTGGCTCAGGAACACTTCAAACTTTCAGAAGG

40

69.9

AE008747

sopB

Pathogenicity

SPI5invasiongeneD

protein

LT2

AAATACACTCACGCATAACGGGCATCACTATACCAACACG

40

69.8

AE008834

sopD

Pathogenicity

Secretedouterprotein

LT2

CCTGCCCGGCTCATCAAGATCTGTTTACTATCAAGATGGA

40

70.2

AE006471

spvC

Pathogenicity

Salm

onel

laplasm

idvirulence:hydrophilic

protein

LT2

GTACAGTGCGTCGTTTCTCCACAAGACACGGCAATTTATA

40

70.0

L16014

hydH

Pathogenicity

Enterotoxin

sensory

kinase

LT2

CTCATTCGCCGTGAATCTCAACTGAATCTCTCTGCTTTGG

40

70.1

AE008826

iroB

Pathogenicity

Putativeglycosyltransferase

LT2

GACCACTGATTGCCGCTAAGTATGACATTCCGGTAGTGAT

40

70.4

AE008719

sfbA

Pathogenicity

Putativeperiplasm

iciron-bindinglipoprotein

LT2

GCAGCATCGACGCCAACCTCTTTCAACATACCCTCTATTT

40

71.1

AE008762

slyA

Pathogenicity

MarR

familytranscriptionalregulatorforhem

olysin

LT2

GCCAAACTTGAACACAATATTATGGAATTGCACTCTCACG

40

67.5

B. Malorny et al. / Molecular and Cellular Probes 21 (2007) 56–65 61

ARTICLE IN PRESSTable

2(c

onti

nued

)

Accessionnumber

Probename

Marker

groupGenefunction

Controlstrain

Probesequence

Length

Meltingtemp(1C)

AF026035

sirA

Pathogenicity

InvasolSirA:regulatorofinvasionproteins

LT2

ATGATGACACATATCCGAATGCCAGTAACAATGCCGAAGC

40

70.3

AE008753

phoQ

Pathogenicity

Sensory

kinase

protein

LT2

ACGCTTGTAAATATTGTCTGGAGTTTGTCGAGATTTCGGC

40

69.0

D14156

wcdA

Pathogenicity

UDP-glucose/G

DPmannose

dehydrogenase

inVipolysaccharidebiosynthesis

RKI-Ty1

ATCAAGCCACTACGATGCGATCATTGTTGCAGTAGGACAT

40

70.7

D14156

wzf

Pathogenicity

Vipolysaccharideexport

inner-m

embraneprotein

RKI-Ty1

CATTAGTTATTAAAGAGCGACGTGTTCAGCAAGCCAAGCA

40

69.3

AE008694

bcfC

Fim

brial

Fim

brialusher,bovinecolonisationfactor

LT2

GGGAAAGTACAACTCGATATCAATCAACAGCTAGGCGGGT

40

70.4

AE008868

lpfD

Fim

brial

Longpolarfimbrialoperonprotein

LT2

CTATGCGATGTCCTGTGAATGCCCTGATGATACCTCTCTT

40

70.1

AE008708

safC

Fim

brial

Putativefimbrialusher

protein

LT2

TGTAAGTGCTAGTTGGCAGATGACTTCACCATCACACGGT

40

71.5

L11008

sefA

Fim

brial

Fim

brialprotein

00-04

TGGTACTCTCAGCATTACTGCTACTGGTCCACATAACTCA

40

69.3

AE008710

stbD

Fim

brial

Fim

bialusher

protein

LT2

TCGGTTTGCCAACGGTGATTAGCGTCAGTAATAGTGAAAC

40

70.0

AE008795

stcC

Fim

brial

Paralputativeoutermem

braneprotein

LT2

TCCGCCAGTCAGACTTACGACGAGGATCATAATGAAGATA

40

69.2

AE008839

stdB

Fim

brial

Putativeoutermem

branefimbrialusher

protein

LT2

CACAGTCCAACAATAACTACATGCTCAGCCTCAACAAGGT

40

69.8

AE008703

stfE

Fim

brial

Putativeminorfimbrialsubunit

LT2

CCTGAGCTGTAACGGCAGAGTGAGCGATTACCTGAAGTTA

40

71.4

AE008702

stiC

Fim

brial


LT2

ATTATCAGTAAGCATCCCGCAACTCTACATCGCCAACAAC

40

69.9

AE008915

stjB

Fim

brial


protein

LT2

ACGGATAATGACAACGACTCTCGCAGTATAATGGCTTCCT

40

70.0

AE008916

STM4595

Fim

brial

Putativefimbrialchaperoneprotein

LT2

AAATTTACGATCAGGCGTGTACGGTTCAGGTGAATGGCTC

40

71.2

AE008826

hin

Serotyping

RegulatorforfljA

:DNA-invertase

hin

LT2

ATTAGTCGGCTATTAGAGAAAGGCCATCCTCGGCAGCAAC

40

71.7

AE008787

fliC

Serotyping

Filamentstructuralprotein,identifies

allH

antigens

LT2

TGTCGCTGTTGACCCAGAATAACCTGAACAAATCCCAGTC

40

70.9

AE008787

fliC_i

Serotyping


iantigen

LT2

GTTGATAAGACGAACGGTGAGGTGACTCTTGCTGGCGGTG

40

73.5

X04505

fliC_r

Serotyping


rantigen

01-2132

ACTACCTTAGGTGGTACTCCTGCTATTACTGGTGATCTGA

40

68.5

AL627272

fliC_d

Serotyping


dantigen

03-1949

TTAGCAAGCGACCTTGACAAACATAACTTCAGAACAGGCG

40

70.3

M84980

fliC_g,m

Serotyping


gcomplexassociatedantigens

00-04

GCGATAGCTGGTGCCATTAAAGGTGGTAAGGAAGGAGATA

40

70.2

U06201

fliC_m,t

Serotyping


m,tantigen

00-102

CGGTAGTAACTGACACCACTGCTCCAACTGTTCCTGATAA

40

70.1

AJ292277

fljB_l,w

Serotyping


l,w

andl,vantigens

01-2132

GTAACCGAAACGCAGCCAAAACCTGTAGCTCTCAGTACAG

40

71.2

AJ292278

fljB_e,n,x

Serotyping

Filamentstructuralprotein

identifies

e,n,x

ande,n,z15antigens

99-601

CTACTGTTACAGGTGATACCGCTGTTACTAAGGTACAGGT

40

68.2

AE008826

fljB_1,2

Serotyping


1,2

1,5,1,6,and1,7

antigens

LT2

ACAATGCCTGCTGGTGCGACAACTAAAACAGAAGTACA

38

70.1

AJ292284

fliC-e,h

Serotyping


e,hantigen

99-929

GGCAAGTACTATGCTGCAACCTATGACGAAGGTACAGGTA

40

70.1

AE008758

wzy

a(1-2)

Serotyping

a-1-2

polymerase:serogroupB

LT2

GCTTTCGGGTATGGTGAACTATACGCAGATTTCGGGCTTT

40

71.1

AE008792

rfbD

Serotyping

TDP-dehydrorhamnose

synthetase:SerogroupA,B,C

2-C

3,D

1,D

2LT2

TCTGCCTCAATGGGAATTAGGAGTTAAGCGTATGCTGACT

40

70.0

AL627273

rfbE(A

,D)

Serotyping

CDP-tyvelose-2-epim

erase:serogroupA

andD

00-04

GGTGTTCAGGCATTCATCAATGTATGGTGGGAGACAGTTT

40

69.9

X60665

orf

9.6

Serotyping

aO-polysaccharidepolymerase:serogroupE1

99-929

TGGTTATGTCGGAGTATTCCTGCATGGTTTAATCTTGGGT

40

69.1

X60666

orf

17.4

Serotyping

bO-polysaccharidepolymerase:serogroupE1andE4

99-929

ATGGTCCGTTCCTGTCTACATTGCATTAGGTTTGCTACTG

40

69.6

AE008792

abe(B)

Serotyping

CDP-abequose

synthase:serogroupB

LT2

ACCTTCATATACTGAGTATCAAGTTGGAACTGGTGCTGGG

40

68.9

X61917

abe(C

2-C

3)

Serotyping

CDP-abequose

synthase:serogroupC2–C3

99-601

TGTCCTATTACCAACAAGACTGCTTGAGTTAATGCCAGCG

40

69.8

AL627273

prt

(A,D

)Serotyping

Paratose

synthase:serogroupsA

andD

00-04

CGACATACTGTGATTGGCTTAGCAAGGAAGAGGAACAATG

40

68.9

X56793

wbaV

(B)

Serotyping

Abequosyltransferase:serogroupB

LT2

GCAGTGATGATGCTCTTGCGAAAGACTCGTTAGCGATATT

40

70.0

X56793

wbaU

(B,D

)Serotyping

Manosyltransferase

(a1-4

linkage):serogroupsBandD1

LT2

CTGATGTTGACGCAATAATCTGTAGCAAGGTACACGCTGA

40

69.8

M84642

wbaA

Serotyping

O-antigen-polymerase:serogroupC1

00-102

ATTACTTGTGCTTGGTGCCATTCTATCATTGCCTTTGTCA

40

69.0

M65054

wbaV

(D1)

Serotyping

Tyvelosyltransferase:serogroupD

00-04

ATCTAATCCACATCGGCGATGGTTAAATGGTGGCAGTAGA

40

70.2

X60665

wbaO

(E1,D

2)Serotyping

Manosyltransferase

(b1-4

linkage):serogroupsE1andD2

99-929

GGCTCCCAGACCTTTACTGAAGTTGATCGGCAAGTGTATA

40

70.2

AF017148

wzy

a(1-6)

Serotyping

a-1-6

polymerase:O27serovarfactor

98-3363

ATGGCAAGTCGTGTCTGGAATATCTCGATGGGATTATCAG

40

69.0

AY334017

wzx

(O6,14)

Serotyping

Oantigen

flippase:O6,14serovarfactor

03-1949

TAAAGCGACCTTGAGTATTGGGCTCACTGCTGTAGTAGTT

40

69.9

AE008737

STM0900

Phages

Fels-1prophageprotein

LT2

GTTTCATTGAGTACGGATTATCGGGTGACGAAATCGGGTA

40

69.3

AE008824

STM2740

Phages


LT2

GAACAGTGGCTCAAAGAGAAGAAACGGACCAGCGATCTTA

40

70.7

AE008823

STM2701

Phages


LT2

CTTAAAGCCGGGAAACTGACCATCGATTATGACTACACGC

40

70.0

AE008819

STM2616

Phages

Gifsy-1

prophageprotein

LT2

CGGCAGGTTGGTCTAGACATGTTCGTTGGAAAAGTAGAG

39

69.8

AE008743

SseI

Phages

Gifsy-2

prophageputativetypeIIIsecreted

protein

LT2

ACTTACAACCTAACCAGTGATATTGATGCTGCGGCCTATC

40

69.4

AF254762

gtgA

Phages

Gifsy-2

prophageprotein

LT2

CAGAACACCAAGATAATCCTTCGCAATTACGCCTCCAACA

40

70.0

AE016848

sopE1

Phages

Translocatedeffectorprotein,encoded

byP2-likebacteriophage

99-601

TAAGAACACTGAGTCTTCTGCAACACACTTTCACCGAGGA

40

70.2

AF200952

sopE2

Phages

Secretedouterprotein,phageremnant

LT2

CTATGCTCGTCAGACATGCGAAGCCATATTATCAGCCGTG

40

71.1

AE008896

STM4210

Phages

Putativemethyl-acceptingchem

otaxisprophageprotein

LT2

GTTGAAGAGGGTATGCGTGAAGCCAAAGAGATGATGGATG

40

70.1

AF378725

cro

Phages

PutativeprophageCro

protein

SUO1

GCCTCATGTGAACGATAGCAGAACCATATTAGCGAAGGTG

40

69.9

AE008720

glxK

Others

Glycerate

kinase

IILT2

CGGAAGTTCTGGCGAACGGTGAACAAAATCTCTACCACAG

40

71.4

AE008784

STM1896

Others

Putativecytoplasm

icprotein

LT2

TTTCAGTAGATGTTTCCGACAATGGTATTTCTGGCGTGGC

40

70.1

AE008911

STM4495

Others

PutativetypeII

restrictionenzyme,

methylase

subunit

LT2

GAAGACTGGCAAGAAGTTGAGGTTATCGGCTGGCTGTATC

40

70.9

B. Malorny et al. / Molecular and Cellular Probes 21 (2007) 56–6562

ARTICLE IN PRESS

Fig. 1. Microarray data of the 19 Salmonella reference strains and the E. coli strain compared to PCR/serotyping results. On the left side the strain number

and serotype is indicated. On the top the terms serotype, fimbriae, pathogenicity, phages, resistance and others designate the marker groups. A column

within each marker group represents one microarray probe. A white box indicates the absence of the target sequence in the microarray and PCR analysis

within the strain. A grey box indicates the presence of the target sequence in the microarray and PCR analysis within the strain. A red box indicates a

discordance between the microarray and PCR detection of the target sequence. The yellow box indicates an uncertain result for the microarray

(normalisation ratio between 0.3 and 0.4).

B. Malorny et al. / Molecular and Cellular Probes 21 (2007) 56–65 63

with respect to different environmental adaptations orselection pressure. The sharing of a similar subset offimbrial clusters between different serotypes (e.g. theserotypes Typhimurium, Enteritidis, Hadar and Saintpaul)might be a prerequisite for the successful worldwide spreadof a serotype.

3.3.3. Pathogenicity-associated determinants

Fourteen probes were designed to identify genes locatedon the five major SPIs [11]. The selection focused on targetgenes whose functions are known and which are uniformlydistributed over the whole islands. All 19 Salmonella

reference strains gave positive signals with 13 probes. TheavrA probe was negative with strains 98-3363 (S.Schleissheim), RKI-Ty1 (S. Typhi) and 02-102 (S. Ora-nienburg). Deletions of the avrA gene in distinct strainshave been previously described [24]. The avrA targetsequence was absent in three strains (in addition, virulencegenes have been selected) which are described to beencoded on the pSLT plasmid of strain LT2 (spv-locus)or elsewhere in the genome (sopD, hydH, iroB, sfbA, slyA,sirA, phoQ, wcdA, wzf). The hydH, sfbA, slyA, and phoQ

genes were present in all Salmonella strains tested, the sopD

gene was absent in strain 01-1543 (S. Paratyphi) and thesirA gene was absent in strains 98-3363 (S. Schleissheim)and RKI-Ty1 (S. Typhi). The sirA gene encodes, togetherwith the barA gene, a two-component sensor kinase and aresponse regulator by increasing the expression of virulencegenes and decreasing the expression of motility genes [25].The lack of the system might play a role in decreasedpathogenicity of Salmonella or host adaptation. The iroB

target sequence was absent in strain SUO8 (S. [4,5,12:i:-]).

3.3.4. Phage-associated genes

Ten phage-associated probes have been selected fromvarious prophages described for S. Typhimurium, such asGifsy-1, Gifsy-2, Fels-1 and Fels-2 and other phages. Fourof these probes identify virulence-associated genes encodedwithin the different phage genome (sopE1, sopE2, sseI,gtgA). The sopE1 probe was positive with DNA of strainsSUO6 (S. Typhimurium), 01-1380 (S. Saintpaul), 99-601(S. Hadar) and 00-4 (S. Enteritidis). The sopE2 gene isencoded by a defective phage haboured by all Salmonella

strains tested. Surprisingly, the prophage Gifsy-2 encodedgenes sseI and gtgA were not uniformly present. Whereasboth genes were found in all S. Typhimurium strains, thesseI gene was only present in strain 98-3363 (S. Schleis-sheim) and 00-4 (S. Enteritidis), and the gtgA gene wasonly present in 99-601 (S. Hadar), and 99-929 (S. Anatum)and 00-4 (S. Enteritidis). These data indicate that theGifsy-2 prophage can be partly present in other serotypes.

3.3.5. Antibiotic resistance markers

Antibiotic resistance genes are linked to phenotypicresistance against certain antibiotics (Table 2). The DNAmicroarray contains 35 probes identifying nine differentphenotypic resistances against antimicrobial agents, mer-cury and quaternary ammonium compound and disin-fectants, as well as integron-carrying strains. These markergenes have been previously well described within Salmo-

nella [26,27]. The agreement of these probes between PCRand DNA microarray signals was 99.4%. There were fourdisagreements, one with the floR probe of strain 00-419(S. Typhimurium), one with the tet(A) probe of strain01-2132 (S. Goldcoast), one with the strA probe of strain

ARTICLE IN PRESSB. Malorny et al. / Molecular and Cellular Probes 21 (2007) 56–6564

SUO8 (S. [4,5,12:i:-]) and one with the int1 (integrase 1)probe of strain 01-1380 (S. Saintpaul). An explanation forthe disagreements between the DNA microarray and PCRmight be that DNA polymorphism within the gene, whichcannot be identified by this probe, exists. This has to beelucidated. For strain 99-601 (S. Hadar) a resistanceagainst ampicillin was only phenotypically identified, butnot on genotypic level (PCR or microarray). Apparently,this resistance is caused by another determinant differentfrom the ones chosen for confirmation.

It was shown that some antibiotic resistance genes can beencoded by very small plasmids of approximately 6 kb [26].An example is strain SUO5, which encodes a sul2 gene on a6 kb plasmid. Its positive microarray signal shows that theDNA purification method is suitable to co-purify suchsmall plasmids along with the chromosomal DNA.

3.3.6. Other determinants

Three probes (glxK, STM1896, STM4495) did not fit tothe other marker groups. The glxK is a gene within theregion encoding the allantoin-glyoxylate pathway. Pre-viously, it was published that this region is devoid of S.[4,5,12:i:-] strains [3]. Our data indicate the absence inSUO1, 00-419 (S. Typhimurium), SUO8 (S. [4,5,12:i:-]),01-2132 (S. Goldcoast) and 03-1949 (S. Lindern).STM1896 and STM4495 have been described as potentialcandidates for S. Typhimurium specific genes [2]. Thiscould not be confirmed since STM1896 was also detected in01-1380 (S. Saintpaul) and 99-929 (S. Anatum). TheSTM4495 probe gave a signal with all S. Typhimuriumstrains and 01-1543 (S. Paratyphi B).

3.3.7. E. coli reference strain

One E. coli strain (EC227) was used as a reference for thetet(D) gene. However, it is not surprising that the strainshowed also strong DNA microarray signals for marT,phoQ, fliC_i, sfbA, sseI, but not in the corresponding PCR.For most of the genes, homologue sequences have beendescribed in E. coli [28]. Several antibiotic resistance genescould also be identified in congruence with the PCR andphenotypic resistance results. The resistances againstchloramphenicol and trimethoprim were only phenotypi-cally identified, but not on the genotypic level (PCR ormicroarray) indicating other responsible resistance DNAdeterminants. The aadB probe indicated the presence of thegentamicin resistance encoding gene, but by PCR andphenotypically, the strain was negative.

3.4. Salmonella reference strain-specific characteristics

In Spain in 1997, a multidrug-resistant fljB-lacking S.

enterica serotype [4,5,12:i:-] emerged. Previously, a micro-array analysis of the genome compared to the LT2 genomeidentified five deleted regions [3]. The microarray resultswith the S. [4,5,12:i:-] strain used in this study (SUO8)confirmed the absence of selected genes from the fivechromosomal regions. Probes specific for glxK, iroB, fljB

([1,2]-antigen), hin, STM900 (Fels-1 prophage), STM2701(Fels-2 prophage), and STM2616 (Gifsy-1 prophage) didnot give a signal. In addition, Guerra et al. [29] identified invarious S. [4,5,12:i:-] strains, a class 1 integron habouringvarious antibiotic resistance genes located on large 140 kbor 120 kb plasmids. The microarray confirmed the anti-biotic resistance genes dfrA12, aadA1, aadA2, blaTEM,aac(3)-IV, cmlA1, sul1, sul2 and tetA and identified, inaddition, a sul3 gene which had not been previouslydescribed for this serotype.Prager et al. [30] proposed that the virulence genes sopE1

and avrA can be used for identifying systemic (SPV) andenteric (EPV) pathovars of S. Paratyphi B. For systemicpathovars, the lack of avrA and the presence of sopE1 werepostulated. Based on the presence or absence of the sopE1,avrA, sopB, sopD and sptP genes, the authors defineddifferent EPV and SPV variants. According to that schemethe Paratyphi B reference strain 01-1543 used in this studycan be classified as an EPV variant 3 (lack of sopD andsopE1, presence of avrA).The low-density microarray described here can be regarded

as a principle of proof for introducing a rapid andcomprehensive genotypic characterisation and typing ofSalmonella isolates using approx. 40mer oligonucleotides asprobes. Such data are necessary in epidemiological andoutbreak studies as well as for risk assessment studies inrespect to the dissemination of antibiotic resistance inmicroorganisms. They allow to estimate virulence and hostspecificity for an individual Salmonella isolate. Currently,traditional serotyping of Salmonella, phenotypic resistancedeterminations and molecular typing methods, such as pulsed-field-gel-electrophoresis (PFGE), need highly experienced staffand a different laboratory equipment. This microarray coversthe most important determinants of all marker groups, suchas pathogenicity, resistance and strain typing, in one testsystem that can be performed within 24h. It is a step towardsa simple test system aiming at maximum genetic straininformation. In a first step, the microarray is interesting forSalmonella Reference Laboratories and hence, especially, forend-user laboratories not equipped with the various analyticalSalmonella-typing techniques. Further developments to im-prove the microarray are in progress. For example, the use ofan internal probe hybridisation control (IHC) indicating anyirregularity during the printing or hybridisation process will beinevitable in the future in order to increase the reliability of thedata. A specific IHC oligonucleotide probe would be mixedwith each target gene probe and printed on the slide. Cy5-labelled Salmonella DNA would indicate the presence of thespecific Salmonella sequence, and Cy3-labelled DNA com-plementary to the IHC sequence would indicate the presenceof the IHC. The success of this concept has been shownpreviously [31]. Another improvement of the microarraywould be the definition of more discriminative thresholdlevels. This aim could be reached by the implementation ofseveral reference probe spots of the ttrC gene distributed overthe microarray in order to overcome possible probe signalgradients. To reduce the cost, the current fluorescence DNA


labelling could be replaced by biotin DNA labelling andsubsequent detection with a streptavidin-gold conjugatecomplex as described by Sachse et al. [32]. Single glass slidesas carrier for the probes could be replaced by platforms with ahigher throughput.

Acknowledgements

The work was supported by the Bundesministerium furErnahrung, Landwirtschaft und Verbraucherschutz(BMELV).

Appendix A. Supplementary data

The online version of this article contains additionalsupplementary data. Please visit doi:10.1016/j.mcp.2006.08.005.

Appendix B. Supplementary data

References

[1] Grimont PAD, Grimont F, Bouvet P. Taxonomy of the genus

Salmonella. In: Wray C, Wray A, editors. Salmonella in domestic

animals. Oxon, UK: CABI Publishing; 2000. p. 1–17.

[2] Porwollik S, Wong RM, McClelland M. Evolutionary genomics of

Salmonella: gene acquisitions revealed by microarray analysis. Proc

Natl Acad Sci USA 2002;99:8956–61.

[3] Garaizar J, Porwollik S, Echeita A, Rementeria A, Herrera S, Wong

RM, et al. DNA microarray-based typing of an atypical monophasic

Salmonella enterica serovar. J Clin Microbiol 2002;40:2074–8.

[4] Porwollik S, Boyd EF, Choy C, Cheng P, Florea L, Proctor E, et al.

Characterization of Salmonella enterica subspecies I genovars by use

of microarrays. J Bacteriol 2004;186:5883–98.

[5] Reen FJ, Boyd EF, Porwollik S, Murphy BP, Gilroy D, Fanning S, et

al. Genomic comparisons of Salmonella enterica serovar Dublin,

Agona, and Typhimurium strains recently isolated from milk filters

and bovine samples from Ireland, using a Salmonella microarray.

Appl Environ Microbiol 2005;71:1616–25.

[6] Boyd EF, Porwollik S, Blackmer F, McClelland M. Differences in

gene content among Salmonella enterica serovar Typhi isolates. J Clin

Microbiol 2003;41:3823–8.

[7] Porwollik S, McClelland M. Lateral gene transfer in Salmonella.

Microbes Infect 2003;5:977–89.

[8] Matsui H, Bacot CM, Garlington WA, Doyle TJ, Roberts S, Gulig

PA. Virulence plasmid-borne spvB and spvC genes can replace the 90-

kilobase plasmid in conferring virulence to Salmonella enterica

serovar Typhimurium in subcutaneously inoculated mice. J Bacteriol

2001;183:4652–8.

[9] Figueroa-Bossi N, Uzzau S, Maloriol D, Bossi L. Variable assortment

of prophages provides a transferable repertoire of pathogenic determi-

nants in Salmonella. Mol Microbiol 2001;39:260–71.

[10] Ehrbar K, Hardt WD. Bacteriophage-encoded type III effectors in

Salmonella enterica subspecies 1 serovar Typhimurium. Infect Genet

Evol 2005;5:1–9.

[11] Kingsley RA, Baumler AJ. Pathogenicity islands and host adaptation of

Salmonella serovars. Curr Top Microbiol Immunol 2002;264:67–87.

[12] McClelland M, Sanderson KE, Spieth J, Clifton SW, Latreille P,

Courtney L, et al. Complete genome sequence of Salmonella enterica

serovar Typhimurium LT2. Nature 2001;413:852–6.

[13] Aarts HJ, Guerra B, Malorny B. Molecular methods for detection of

antibiotic resistance. In: Aarestrup FM, editor. Antimicrobial

resistance in bacteria of animal origin. Washingthon, DC: ASM

Press; 2005. p. 37–45.

[14] Carattoli A. Plasmid-mediated antimicrobial resistance in Salmonella

enterica. Curr Issues Mol Biol 2003;5:113–22.

[15] Kane MD, Jatkoe TA, Stumpf CR, Lu J, Thomas JD, Madore SJ.

Assessment of the sensitivity and specificity of oligonucleotide

(50mer) microarrays. Nucleic Acids Res 2000;28:4552–7.

[16] Malorny B, Paccassoni E, Fach P, Bunge C, Martin A, Helmuth R.

Diagnostic real-time PCR for the detection of Salmonella in food.

Appl Environ Microbiol 2004;70:7046–52.

[17] Hughes TR, Mao M, Jones AR, Burchard J, Marton MJ, Shannon

KW, et al. Expression profiling using microarrays fabricated by an

ink-jet oligonucleotide synthesizer. Nat Biotechnol 2001;19:

342–7.

[18] Faucher SP, Porwollik S, Dozois CM, McClelland M, Daigle F.

Transcriptome of Salmonella enterica serovar Typhi within macro-

phages revealed through the selective capture of transcribed

sequences. Proc Natl Acad Sci USA 2006;103:1906–11.

[19] Chen S, Zhao S, McDermott PF, Schroeder CM, White DG, Meng J.

A DNA microarray for identification of virulence and antimicrobial

resistance genes in Salmonella serovars and Escherichia coli. Mol Cell

Probes 2005;19:195–201.

[20] Pelludat C, Prager R, Tschape H, Rabsch W, Schuchhardt J, Hardt

WD. Pilot study to evaluate microarray hybridization as a tool for

Salmonella enterica serovar Typhimurium strain differentiation.

J Clin Microbiol 2005;43:4092–106.

[21] van Hoek AH, Scholtens IM, Cloeckaert A, Aarts HJ. Detection of

antibiotic resistance genes in different Salmonella serovars by oligonu-

cleotide microarray analysis. J Microbiol Methods 2005;62:13–23.

[22] McQuiston JR, Parrenas R, Ortiz-Rivera M, Gheesling L, Brenner F,

Fields PI. Sequencing and comparative analysis of flagellin genes

fliC, fljB, and flpA from Salmonella. J Clin Microbiol 2004;42:

1923–32.

[23] Reeves P. Evolution of Salmonella O antigen variation by inter-

specific gene transfer on a large scale. Trends Genet 1993;9:17–22.

[24] Amavisit P, Lightfoot D, Browning GF, Markham PF. Variation

between pathogenic serovars within Salmonella pathogenicity islands.

J Bacteriol 2003;185:3624–35.

[25] Goodier RI, Ahmer BM. SirA orthologs affect both motility and

virulence. J Bacteriol 2001;183:2249–58.

[26] Guerra B, Junker E, Miko A, Helmuth R, Mendoza MC.

Characterization and localization of drug resistance determinants in

multidrug-resistant, integron-carrying Salmonella enterica serotype

Typhimurium strains. Microb Drug Resist 2004;10:83–91.

[27] Miko A, Guerra B, Schroeter A, Dorn C, Helmuth R. Molecular

charcaterization of multiresistent d-tartrate-positive Salmonella enterica

serovar Paratyphi B isolates. J Clin Microbiol 2002;40:3184–91.

[28] McClelland M, Florea L, Sanderson K, Clifton SW, Parkhill J,

Churcher C, et al. Comparison of the Escherichia coli K-12 genome

with sampled genomes of a Klebsiella pneumoniae and three

Salmonella enterica serovars, Typhimurium, Typhi and Paratyphi.

Nucleic Acids Res 2000;28:4974–86.

[29] Guerra B, Soto SM, Arguelles JM, Mendoza MC. Multidrug

resistance is mediated by large plasmids carrying a class 1 integron

in the emergent Salmonella enterica serotype [4,5,12:i:-]. Antimicrob

Agents Chemother 2001;45:1305–8.

[30] Prager R, Rabsch W, Streckel W, Voigt W, Tietze E, Tschape H.

Molecular properties of Salmonella enterica serotype Paratyphi B

distinguish between its systemic and its enteric pathovars. J Clin

Microbiol 2003;41:4270–8.

[31] Tomioka K, Peredelchuk M, Zhu X, Arena R, Volokhov D,

Selvapandiyan A, et al. A multiplex polymerase chain reaction

microarray assay to detect bioterror pathogens in blood. J Mol Diagn

2005;7:486–94.

[32] Sachse K, Hotzel H, Slickers P, Ellinger T, Ehricht R. DNA

microarray-based detection and identification of Chlamydia and

Chlamydophila spp. Mol Cell Probes 2005;19:41–50.

dx.doi.org/10.1016/j.mcp.2006.08.005

dx.doi.org/10.1016/j.mcp.2006.08.005

Molecular characterisation of Salmonella strains by an oligonucleotide multiprobe microarray

Documents