using antimicrobial agents in preventative medicineprogrammes Multidrug resistant Salmonella enterica isolated from conventional pigfarms The Veterinary Journal

(1)

Original Article

Multidrug resistant Salmonella enterica isolated from conventional pig farms using antimicrobial agents in preventative medicine

programmes

Karla Cameron-Veas

^a

, Lorenzo Fraile

^b,c

, Sebastian Napp

^a

, Victoria Garrido

^d

, María Jesús Grilló

^d

, Lourdes Migura-Garcia

^a,

*

aInstitutdeRecercaiTecnologiaAgroalimentàries,CentredeRecercaenSanitatAnimal,UniversitatAutònomadeBarcelona,Spain

bDepartamentodeProducciónAnimal,UniversidaddeLleida,Lleida,Spain

cAgrotecnioCenter,Lleida,UniversidaddeLleida,Spain

dInstitutodeAgrobiotecnología,ConsejodeGobierno,GobiernodeNavarra,UniversidadPúblicadeNavarra,Mutilva,Navarra,Spain

ARTICLE INFO

Keywords:

Ceftiofur b^-Lactams Multidrugresistance Pig

Salmonellaenterica

ABSTRACT

Alongitudinalstudywasconductedtoinvestigatethepresenceofmultidrugantimicrobialresistance (multi-AR)inSalmonellaentericainpigsrearedunderconventionalpreventativemedicineprogrammes inSpainandthepossibleassociationofmulti-ARwithceftiofurortulathromycintreatmentduringthe pre-weaningperiod.Groupsof7-day-oldpigletsweretreatedbyintramuscularinjectionwithceftiofur onfourfarms(n=40pigletsperfarm)andwithtulathromycinonanotherfourfarms(n=40pigletsper farm).Acontrolgroupofuntreatedpiglets(n=30perfarm)waspresentoneachfarm.Faecalswabswere collectedforS.entericaculturepriortotreatment,at2,7and180dayspost-treatment,andatslaughter.

Minimal inhibitory concentrations of14 antimicrobial agents, pulsed-field gel electrophoresisand detectionofresistancegenesrepresentingfivefamiliesofantimicrobialagentswereperformed.Plasmids carrying cephalosporin resistant(CR) genes werecharacterised. Sixty-sixS. enterica isolates were recoveredfromfiveofeightfarms.Forty-sevenisolatesweremulti-ARandfourcontainedblaCTX-Mgenes harbouredinconjugativeplasmidsoftheIncI1family;threeoftheseisolateswererecoveredbefore treatmentwithceftiofur.ThemostfrequentARgenesdetectedweretet(A)(51/66,77%),sul1(17/66,26%);

tet(B)(15/66,23%)andqnrB(10/66, 15%).Adirectrelationbetweentheuseofceftiofurintheseconditions andtheoccurrenceofCRS.entericawasnotestablished.However,multi-ARwascommon,especiallyfor ampicillin, streptomycin, sulphonamidesand tetracycline. These antibioticsare used frequently in veterinarymedicineinSpainand,therefore,shouldbeusedsparinglytominimisethespreadofmulti-AR.

Introduction

Salmonella enterica is a major foodborne pathogen, causing infections in human beings and animals worldwide. In 2014, 89,873conﬁrmedhumancasesofsalmonellosiswerereportedin theEuropeanUnion(EU)(EFSA,2017).ControlprogrammesforS.

entericaineggsandpoultryhavebeeneffectiveinreducingthe number of salmonellosis cases in human beings (EFSA, 2014).

However,thistrendislevellingout,owingtothepersistenceofS.

entericainpigs andporcineproducts(Pireset al.,2011).Onthe basisofexaminationofmesentericlymphnodesoffatteningpigs, thereisawiderange(0–36.4%)inthefrequencyofdetectionofS.

entericabetweencountriesintheEU(EFSA,2015).InSpain,amajor producerofpigsintheEU,S.entericawasdetectedin30%ofpigs (EFSA,2015).

Vaccinationof pigsagainst S.entericaserovarsthatarenon- pathogenic for pigs is not recommended, since pigs can be asymptomaticallyinfectedandcantransmitthebacteriatohuman beings(SanRománetal.,2013).S.entericacontrolprogrammesin pigs are mainly based on hygiene/disinfection, biosecurity measures and farm management practices. Treatment with antimicrobialagentsarenecessaryforcontrolofclinicaloutbreaks involvingbacteriaasprimaryorsecondary pathogens,andhave beenusedasametaphylacticandprophylactictoolwhenthereisa highprobabilityof anoutbreak of unknownbacterialaetiology (Barton, 2014).The selective pressure exerted by antimicrobial agents may contribute to the emergence of bacteria with antimicrobial resistance (AR) (Garcia-Migura et al., 2014). Pigs

*Correspondingauthor.

E-mailaddress:[email protected](L.Migura-Garcia).

https://doi.org/10.1016/j.tvjl.2018.02.002

ContentslistsavailableatScienceDirect

The Veterinary Journal

j o u r n al h o m e p a g e : w w w . el s e v i e r . c o m / l o c a t e / t v j l

(2)

carryingARbacteriamayhaveimplicationsforpublichealth,since thereisariskoffoodbornetransmissiontoconsumersthroughthe foodchain.

Thirdandfourthgenerationcephalosporins,suchasceftiofur and cefquinome, are licensed for the treatment of systemic bacterial infections in pigs (Cameron-Veas et al., 2015). These

b

^-lactamantimicrobialagents are someof themostimportant compounds used in human medicine, constituting the main therapeuticchoice for treatment of infections caused byEnter- obacteriaceae (Collignonet al., 2009).The possible selection of cephalosporinresistant (CR) S.enterica, togetherwithconcerns relatingtotheirentryintothefoodchain, hasraisedquestions regardingtheuseoftheseantimicrobialagentsinpigs.InSpain,

multi-ARwasdetectedin55%ofS.entericaisolatesin2015(EFSA, 2017).

This longitudinal study was undertaken to evaluate the presence of multi-AR S. enterica on conventional pig farms in Spainthatuseantimicrobialagentsintheirpreventativemedicine programmes.

b

^-Lactam antimicrobial agents (penicillins and cephalosporins) andmacrolides(tulathromycinand tildipirosin) are the most commonly prescribed drugs in pigs during the sucklingperiodinSpain(Moreno,2014).Theaimofthisstudywas toassesstheeffectofceftiofurandtulathromycintreatmentonthe emergenceofCRS.entericaduringthepre-weaningperiod,andin theS.entericapopulationinpigsfromday7untilslaughter.The genotypic and phenotypic diversity of the S. enterica serovars obtained from each of the farms were analysed. Since Table1

Salmonellaentericaserovarsisolatedfromfaecesatdifferentintervalsandtreatmentsadministeredineachfarmduringtherearingcycle.

Farm Antimicrobialagentsusedatdifferentpost-birthintervals Samplingatdifferentpost-birthintervals Numberofisolates/numberofsamplesanalysed(%) Numberofserovars

Intramuscular(day7) Pre-starter/starter1(days21–49) Starter2(days50–79) Day7 Day9 Day14 Day187

1 Untreated Amoxicillin

Colistinsulphate

NA 2/30(7%)

2Rissen

3/30(10%) 3Rissen

3/28(11%) 3Rissen

2/28(7%) 2Rissen

Tulathromycin 0/40 0/40 0/39 2/39(5%)

2Derby

Sows 0/7

Apramycin Tiamulin Oxytetracycline

Tiamulin Oxytetracycline

1/30(3%) 1Panama

3/30(10%) 3Typhimurium

0/30 2/24(8.3%) 2Panama

Tulathromycin 2/40(5%)

2Panama

1/40(2.5%) 1Panama

0/40 3/36(8.3%) 1Panama 2Typhimurium

Sows 1/7(14.2%)

1Typhimurium

6 Untreated 0/30 0/30 0/30 0/24

Tulathromycin 0/40 0/40 0/40 2/34(5.8%)

1Rissen 1Typhimurium

Sows 0/7

7 Untreated 0/30 0/30 0/30 0/16

Tulathromycin 0/40 0/40 0/40 0/30

Sows 0/7

Apramycin Tiamulin Oxytetracycline

Tiamulin Oxytetracycline

13/30(43%) 2Brandenburg

11Rissen^a

4/30(13%) 4Rissen^a

1/30(3.3%) 1Rissen

1/22(4.5%) 1Typhimurium

Ceftiofur 7/40(18%)

3Anatum^a 3Rissen^a 1Brandenburg

5/40(13%) 3Brandenburg

2Rissen

1/38 1Rissen

1/38(2.6%) 1Typhimurium

Sows 1/7(14.2%)

1Brandenburg

4 Untreated 0/30 0/30 0/30 0/18

Ceftiofur 0/40 0/40 0/40 0/35

Sows 0/7

5 Untreated 0/30 0/30 0/26 4/17(24%)

4Rissen

Ceftiofur 0/40 0/40 0/37 1/33(3%)

1Rissen

Sows 0/7

8 Untreated 0/29 0/29 0/21 0/21

Ceftiofur 0/37 0/36 0/34 0/0

Sows 0/7

Totalstrains66 27/614(4.4%) 16/555(0.9%) 5/533(0.9%) 18/415(4.3%)

aAtotaloffourcephalosporinresistant(CR)Salmonellaentericaserovarswerefound.OneS.Anatumisolateswereobtainedatday7inthecephalosporintreatedgroup.

ThreeS.Rissenisolateswereobtainedintheuntreatedgroup(2strainsatday7and1strainatday9).NA,notapplicable.

(3)

cephalosporinsandﬂuoroquinolonesaretheantimicrobialagents ofchoicefortreatmentofhumansalmonellosis,thepresenceof genescodingfortheseantimicrobialagentswasassessed.

Materialsandmethods Studydesign

ThisstudywasapprovedbytheAnimalCareandEthicsCommitteeofthe UniversityofLleida,Spain(approvalnumberDAAM7684;dateofapproval12

October2012).ThiswasalongitudinalstudyonS.entericaprevalencecarriedout onacohortofeightfarmslocatedinCatalonia,Spain.Samplingwascarriedoutat thesametimeasastudyonEscherichiacolipublishedbyCameron-Veasetal.

(2016).Noneofthefarmshadahistoryofcephalosporinuseforatleast2years priortothestudy.ThesamplingperiodstartedinNovember2012andﬁnishedin May2014.Seventy7-day-oldpiglets(sevenseparatelittersperfarm)wereear taggedinbothearsforidentiﬁcationthroughoutthestudyperiod.Thissamplesize wasbasedonanestimatedprevalenceofS.enterica sheddersofatleast40%

(Casanova-Higesetal.,2017).Anincrementofatleast30%ofanimalsexcretingAR bacteriaaftertreatmentwasconsideredtobesubstantive.Onthebasisofthese numbers,thesamplesizeselectedforthestudywas32animalsineachofthe Fig.1.DendrogramillustratingXbaI-pulsefieldgelelectrophoresis(PFGE)profilesandantimicrobialresistancephenotypeofSalmonellaenterica(n=66)isolatedfromfaeces offatteningpigsinCatalonia,Spain.Blacksquaresrepresentresistance.A,Ampicillin(wildtype,WT,8mg/L);C,chloramphenicol(WT16mg/L);S,streptomycin (WT16mg/L);Su,sulphamethoxazole(WT64mg/L);T,tetracycline(WT8mg/L);Ctx,cefotaxime(WT0.25mg/L);Caz,ceftazidime(WT0.5mg/L);Ci,ciprofloxacin (WT0.064mg/L);Na,nalidixicacid(WT16mg/L);G,gentamicin(WT2mg/L);K,kanamycin(WT8mg/L);F,florfenicol(WT16mg/L);Cs,colistin(WT2mg/L);

Tm,trimethoprim(WT2mg/L).Codesofisolates(IdLab)werebasedonthenumbersassignedtothefarm(G),visitnumber(V)andpignumber(C).

(4)

controlandantimicrobialtreatedgroups,withaconﬁdencelevelof95%anda powerof80%.¹

Farmswererandomlydividedintotwogroupsoffourfarmseach,according to the antimicrobial treatmentadministered for experimental purposes, i.e.

ceftiofur or tulathromycin. Within each farm, one group of 7-day-old pigs (n=40)wastreatedwitheitherceftiofur(Naxcel,Zoetis;5mg/kgofbodyweight) ortulathromycin(Draxxin,Zoetis;2.5mg/kgofbodyweight),administeredbya single intramuscular injection (Cameron-Veas et al., 2016). The remaining selectedpigsfromeachfarm(n=30)werekeptasanuntreatedcontrolgroup.

Inmostcases,thegroups remainedseparatedindistantpensoverthestudy period,includingduringtransportationtotheﬁnishingfarm;theexceptionwas farm2,whichwasmanagedunderafarrow-to-ﬁnishcyclesystem.Mostfarms belongedtothesamelargeintegrationsystem;theexceptionwasfarm1,which wasownedbyanindependentfarmer.Duringtherearingperiod,allpigswerefed underastandardnutritionalprogramme(seeAppendix:SupplementaryTable1) set bythe companies,includingthe prophylacticuseof antimicrobialagents (Table1),fromdays21(weaning)to49(pre-starterandstarter1feeds),aswell asfromdays50to70(starter2feed).Ofthe560animalsselectedforinclusionin thestudy,164werenotsampledatsomepointduetoeitherdeathorlossofear tags(Table1).

Faecalsampleswerecollectedfrom7-day-oldpiglets(day7)beforetreatment withtulathromycinorceftiofur,asdetailedinTable1,atdays9and14oflife,and immediatelybeforetheanimalsweresentforslaughter(day1872days).Onday7, faecalsampleswerealsocollectedfromtherespectivesows(n=7perfarm).

Salmonellaentericacultureandcharacterisation

Faecalsamplesweretakenfromtherectumofeachpigandtransportedtothe laboratoryat4Conthesamedayofsampling.Onreceipt,5goffaecesfromeach animalwerehomogenisedin25mLbufferedpeptonewaterandincubatedat37C for24h.Afterincubation,0.1mLhomogenatewereinoculatedontoRappaport- Vassiliadissemisolidmediumandincubatedat42Cfor24and48h(thelatterif negativeat24h),followedbystreaking1mLonXLT4mediumwithandwithout ceftriaxone(1mg/L).CulturemediawerefromMerckandantibioticswerefrom Sigma–Aldrich.OneS.entericaisolatefromeachpositivesamplewasselectedfor confirmation and serotyping at the Laboratori Agroalimentari, Departament d’Agricultura,Ramaderia,Pesca,AlimentacióiMediNatural, Cabrils,Spain,by thereferenceslideagglutinationmethod,forspecificsomaticflagellaandcapsular antigens,usingtheKauffmann–Whitescheme(Popoffetal.,2004).

Antimicrobialsusceptibilitytesting

Minimalinhibitoryconcentrations (MICs)of14 antimicrobialagentswere determinedusing the broth microdilution method (VetMICGN-mo, Swedish NationalVeterinaryInstitute)usingampicillin, chloramphenicol,streptomycin, sulphamethoxazole,tetracycline,cefotaxime,ceftazidime,ciprofloxacin,nalidixic acid,gentamicin,kanamycin,florfenicol,colistinandtrimethoprim(Fig.1)(Sola- Ginesetal.,2015).Epidemiologicalcut-offvaluesweredeterminedaccordingto recommendationsof the EuropeanCommitteeonAntimicrobialSusceptibility Testing(EUCAST).²Multidrugresistancewasdefinedasresistancetoantibioticsof atleastthreedifferentfamilies(Schwarzetal.,2010).

Twodoubledisccombinations(cefotaximewithcefotaxime/clavulanicacidand ceftazidimewithceftazidime/clavulanicacid)wereusedtoconfirmtheextended spectrumb^-lactamase^(ESBL)^phenotype.^Synergy^was^definedâsânîncreaseⁱⁿ zone diameter5mm. Cefoxitin was used for the detection of ampC-type b-lactamase(CLSI,2008).

Antimicrobialresistantgeneticdeterminants

ResistancetocephalosporinswasassessedbyPCRforblaTEM,blaCTXM,blaCMY-1, blaCMY-2andblaSHV(Hasmanetal.,2005).Sequenceanalysiswasperformedusing VectorNTI advance 11(InforMax).The ampliﬁed nucleotide sequenceswere comparedtothoseintheNationalCenterforBiotechnologyInformation(NCBI)data baseusingBLAST.³

Isolatesexhibiting lowsusceptibility tociproﬂoxacinweretested forthe plasmid-mediatedquinoloneresistancegenes(PMQRs)qnrA,qnrB,qnrC,qnrDand qnrS(Wasyl,2014).ThepresenceofresistancegenesfortheaminoglycosidesaadA, strA/strB,aac(3)IV,aadB,aphA1andaphA2,thetetracyclinestet(A),tet(B)andtetC), andthesulphonamidessul1,sul2andsul3werealsotestedbyPCR(Kozaketal., 2009).

Pulsedﬁeldgelelectrophoresisandplasmidcharacterisation

Toassesstheclonalityofthestrainsandtheepidemiologicalrelatednesswithin farms, XbaI pulsed ﬁeld gel electrophoresis (PFGE) macro-restriction was performedusingaChef-DRIISystem(Biorad)(Ribotetal.,2006).PFGEproﬁles were comparedusingFingerprintingII Informatixesoftware (AppliedMaths).

Isolateswereconsideredtohaveauniquepatternwhenatleastonebanddifference wasdetected(Liebanaetal.,2002).BandswereanalysedusingtheDicecoefﬁcient andunweightedpairgroupmethodwitharithmeticaverages(optimisation1.5%

andpositiontolerance1.5%).

RepliconsfromplasmidswithanESBLphenotypewerecharacterisedbyPCR- basedreplicontyping(Carattolietal.,2005).Filtermatingexperimentsusingthe rifampicinresistantE.colistrainHB101,togetherwithplasmidDNAextractionand electroporation, wereperformedasdescribed by Cameron-Veaset al. (2015).

Transconjugants were selected on Luria–Bertani (LB) agar plates containing rifampicin(100mg/L)andceftriaxone(1mg/L).Thepresenceofauniqueplasmid ineachtransconjugantharbouringthecephalosporinresistantgenewasconﬁrmed andtheirsizesweredeterminedusingS1-nucleasedigestion,followedbyPFGE (Bartonetal.,1995).

Statisticalanalysis

McNemar’stest(pairedsamples)wasusedtoevaluatewhethertheprevalence ofS.entericaandCRS.entericainthedifferentherdsdifferedbetweenconsecutive samplingtimesforagiventreatment(ceftiofurortulathromycin)(Kirkwoodand Sterne,2003).Thex²testwasusedtoassesswhethertheprevalenceofS.enterica differedamongherdsortreatmentsatagivensamplingtime.Whenthenumberof observationsforoneofthecategorieswassmall(i.e.<5),Fisher’sexacttestwas used(KirkwoodandSterne,2003).StatisticalanalyseswerecarriedoutusingSAS version9.1.3(IBM)andthesigniﬁcancelevelwassetatP<0.05.

Results

OccurrenceofSalmonellaenterica

Atotalof2117faecalsampleswerecollectedduringthecourse ofthestudy.Sixty-six(3.1%)sampleswerepositiveforS.enterica;

these wereobtained fromfive out of the eight farmssampled (Table1).Onthefivepositivefarms,27isolateswereobtainedfrom day7,16fromday9,fivefromday14and18fromday1872.CRS.

enterica was detected on one farm (farm 2; Table 1), prior to ceftiofuradministration(3/21isolates)andatday9(1/9isolates).

NosigniﬁcantdifferenceswereobservedintheprevalenceofCRS.

entericabetweenconsecutivesamplingtimesaftertreatmentwith ceftiofur.

TheprevalenceofS.entericaatday7washigher(P<0.01)on farm2(20/70,28.6%)thanontheremainingfarms(2/70,2.8%).On farm 2, prior to antimicrobialtreatment, theuntreated control grouphadahigherprevalenceofS.enterica(13/30,43%)thanthat oftheceftiofurtreatedgroup(7/40,17.5%;P=0.018).Thereafter, the prevalence of S. enterica in the untreated group reduced significantly (P<0.01) to 13% of shedders at day 9 in both untreated and treated groups (4/30 and 5/40, respectively) (Table1).Asignificantreduction(P<0.05)inS.entericashedding wasobservedfromdays7to14forboththeuntreatedandtreated groups.Onfarm5,significantdifferences(P=0.022)inS.enterica prevalence were found, at day 187,between pigs treated with ceftiofur(1/33,3.0%)anduntreatedpigs(4/17,24%)(Table1).S.

entericawasisolatedfromsowson2/8farms,aswellasintheir offspring;serovarswerethesamebetweensowsandpigletson farm2,butdifferentonfarm3.

Amongthe66isolatesofS.enterica,themostprevalentserovar was Rissen(38/66,58%),followedbythemonophasicvariantof Typhimurium(9/66,14%),thenPanamaand Brandenburg(7/66, 11%each)(Table1).Theﬁrsttwowerethemostwidelydistributed serovars,sincetheywerepresentinfourandthreeofthepositive farms,respectively.

1 See:http://epitools.ausvet.com.au/content.php?page=home(accessed18Jan- uary2017).

2 See:http://www.eucast.org/(accessed18January2017).

3 See:www.ncbi.nlm.nih.gov(accessed18January2017).

(5)

Phenotypicandgenotypicresistance

AsshowninFig.1,S.entericaisolatesexhibitedresistance to tetracycline (98%), ampicillin (58%), sulphamethoxazole (53%), streptomycin (58%), ciprofloxacin (50%), trimethoprim (32%), gentamicin (20%), chloramphenicol (20%) and nalidixic acid (15%).Twelvepercentoftheisolateswereresistanttoflorfenicol and4%wereresistanttokanamicin.Noneoftheisolatesexhibited resistancetocolistin.Forty-sevenisolatesweremultidrugresis- tant,beingresistanttoatleast3/14antimicrobialagentstested, and exhibiting 19 different antimicrobial resistance profiles.

Among these isolates, 38/66 (57.6%) were resistant to at least fourantimicrobial classes, with31/38 (81.6%) isolates resistant simultaneouslytoampicillin(A),streptomycin(S),sulphonamides (Su) and tetracycline (T) alone or in combination with other antimicrobialagents,thusexhibitingtheASSuTmultidrugresis- tant proﬁle (Table 2). In the case of the S. Typhimurium monophasicstain,ASSuT is associatedwitha resistance region localisedonthebacterialchromosome.Furthermore,16/18(89%) isolates recovered before the animals were slaughtered were multi-resistant. Only one isolate (serovar Rissen) was pan- susceptible (Fig. 1). Four isolates (three S. Rissen and one S.

Anatumfromfarm2)wereresistanttocephalosporins(Table2).

Themostprevalentantimicrobialgenotypesweretet(A)(77%), sul1(27%)andtet(B)(23%)(Table2).TheqnrBgenewasdetectedin 10(15%)isolatesamongfourserovars,namelyBrandenburg(n=5), AnatumandRissen(n=2each),andPanama(n=1)(Table2);nine of theseisolates wererecovered from thesame farm (farm2), whileS.PanamawasrecoveredfromFarm3(Fig.1).

Resistance to cephalosporin in isolates from farm 2 was associatedwiththepresenceofblaCTX-M-1and blaCTX-M-14genes intwodifferentS.Rissenisolates,andbla_CTX-M-1inoneisolateofS.

Anatumrecovered fromthree7-dayoldpiglets.AnadditionalS.

RissenblaCTX-M-1wasisolatedfromthesameanimalinthesecond visit(atday9);thisanimalbelongedtothecontrolgroup(Table1).

CRS.entericawasnotdetectedduringfurthervisitstothisfarm.All isolatesharbouringCTX-Mtransferredthegenestotherecipient strainbyconjugationandtransformation.Inallcases,CTX-Mgenes wereharbouredina95kilobaseplasmidbelongingtotheIncI1 incompatibilitygroup.

Macro-restrictionanalysisusing XbaIproduced 12–16bands anddistributedthe66isolatesintoeightmajorclustersconsisting of isolates withrelated PFGE proﬁles (80% identity) and three uniquePFGEpatternsrepresentedbyasingleisolateeach(Fig.1).

PFGE analysis demonstrated highclonality between S. enterica isolates of the same serovar within farms. Indistinguishable

Table2

PhenotypicandgenotypiccharacterisationofantimicrobialresistanceproﬁlesoftheSalmonellaentericaisolates(numbersincurvedbrackets)obtainedinthisstudy,grouped byserovars.

Serovars Phenotype^a(numberofisolates) Genotype(numberofisolates)

Rissen(38) ACSSuTCiNaGTm(1) tet(A)(1)

ASSuTTm(5) tet(A)/sul1(2);aadA/tet(A)/sul1(2);aadA/aac(3)IV/tet(A)/sul1(1)

ASTTm(1) tet(A)(1)

ATCtxCazCiNa(1) tet(A)/blaCTX-M(1)

ATCtxCazCi(1) tet(A)/blaCTX-M(1)

ATCi(2) tet(A)(2)

ATCtxCazCiNaTm(1) tet(A)/blaCTX-M(1)

TCiNa(1) tet(A)(1)

STCi(4) tet(A)(4)

TCi(9) tet(A)/qnrB(2);tet(A)(7)

TCCi(1) tet(A)(1)

TCiF(1) tet(A)(1)

TGF(1) tet(A)(1)

T(7) tet(A)(6);tet(A)/sul1(1)

SuT(1) tet(A)(1)

Pan-susceptible(1) Notfound

Typhimuriummonophasic(9) ASSuTCi(1) tet(B)/sul2(1)

ASSuTG(1) tet(B)/sul2(1)

ACSSuTGTm(1) tet(B)/su1/sul2l(1)

ASuTG(1) tet(A)(1)

ASSuTTm(1) tet(B)(1)

ASSuT(4) tet(B)/sul2(2);tet(B)(2)

Brandenburg(7) ACSSuTCiNaGFTm(5) aadA/strA/B/tet(A)/sul1/qnrB(1);tet(A)/sul1/sul3/qnrB(3);aadA/strA/B/tet(A)/sul1/sul2/qnrB(1)

ACSSuT(1) tet(A)/sul1(1)

CSSuTTm(1) tet(A)/sul1(1)

Panama(7) ASSuTCi(1) tet(B)(1)

ASSuTCiTm(1) tet(B)/qnrB(1)

ASSuTGK(2) aphA1/tet(B)(2)

ASSuT(2) tet(B)(2)

ASSuTK(1) tet(B)(1)

Anatum(3) CSSuTCiTm(1) tet(A)/sul1(1)

ACSSuTCiNaGFTm(1) aadA/strA/B/tet(A)/sul1/qnrB(1)

ACSSuTCtxCazCiTm(1) tet(A)/sul1/qnrB/blaCTX-M(1)

Derby(2) ASSuTTm(1) aadA/tet(A)/sul1(1)

ASSuT(1) aadA/tet(A)/sul1(1)

Ctx,cefotaximeandceftazidime;aminoglycosides:aadA,strA/strB,aac(3)IV,aadB,aphA1,aphA2;tetracycline:tet(A),tet(B),tet(C);sulphonamides:sul1,sul2,sul3;

quinolones:qnrA,qnrB,qnrB,qnrD,qnrS;b^-lactams:^bla^CTX-M^,^bla^CMY^,^bla^SHV^.

aSeeFig.1forantimicrobialnomenclature.

(6)

ﬁngerprintswerepresentinisolatesfromdifferentanimalsand fromisolatesobtainedatdifferentsamplingtimes,indicatingthe persistence of clones over time. Two of the S. Rissen isolates resistanttocephalosporinrecoveredfromthefarm2hadidentical PFGEpatterns(E3G2V1C7RandE3G2V1C50R;Fig.1).

Discussion

A common practice in preventative medicine programmes amonglargepigproducersofSpainistoinject7-dayoldpiglets with one intramuscular dose of ceftiofur during the lactation period.Inthelongitudinalstudycarriedoutonconventionalfarms, wefoundnoevidenceofadirecteffectfromthispracticeonthe emergence of CR S. enterica strains. This conclusion could be inﬂuencedbythelowprevalenceofS.entericaandthelownumber ofCRstrains foundinthefarmsstudied,aswell asbythewide varietyofserovarsisolated,sincedifferentserovarsshowdifferent abilitiestoacquireresistancegenes(Aarestrup,2004).

Inthefour CRisolates recoveredfromfarm 2,CTX-M genes wereharbouredonaconjugativeplasmidof95kb,belongingtothe IncI1 familyof replicons (Rodriguezet al., 2009; Zurﬂuh etal., 2014).Farm2alsohadahighprevalenceofCRE.coliduringthe study period associated with the same group of plasmids (Cameron-Veasetal.,2016).Thepresenceof theseCRstrainsin both treated and untreated cephalosporin groups suggests an exchange of mobile genetic elements between the different Enterobacteriaceae could have occurred in this farm. Moreover, we found that most of S. enterica strains (especially those recoveredonthe last visit tothe farm and beforetheanimals weresent)exhibitedmulti-ARandbelongedtoS.entericaserovars frequently associated withhumaninfections (EFSA, 2011). This ﬁndinghighlights theimportanceofhumaninfections frompig productsifthesemulti-ARstrainsenterthefoodchain.

Althoughgastroenteritiscausedbynon-typhoidalS.entericais mostly self-limiting and treatmentis not required, 5%of the individualswilldevelopbacteraemia,whichispotentiallyfataland requiresantibiotictreatment(Anjumetal.,2011).Inthesecases, the recommended drugs of choice are ﬂuoroquinolones and cephalosporins. Nowadays, the treatment of many infectious diseases, in both humans and animals, relies upon just one or twodrugs (Woolhouseand Farrar, 2014).Arestrictivepolicyof antibioticusageshouldbeimplementedinlivestockintheEU,in ordertoprotecthumanhealth.

ThequinoloneresistancegeneqnrBwasdetectedindifferentS.

enterica serotypes coexisting in the same farm. Resistance to tetracyclinewasfrequentandwascorrelatedwiththepresenceof tet(A)andtet(B)genes.Additionally,resistancetoaminoglycosides encoded by aphA1, aadA and aac(3)IV was also present. As demonstratedin Table 1,at leastfour families of antimicrobial agents (

b

^-lactams, polymyxins, aminoglycosides and tetracy- clines)wereadministeredduringtherearingcycleonallof the conventionalfarmsincluded inthis study, asusuallyoccurs on intensivepigfarms(Moreno,2014;EMA,2015⁴).Inthisstudy,a high number of isolates exhibited a multi-AR phenotype to streptomycin, ampicillinandsulphonamides, which are antimicrobial agents commonly use in veterinary medicine. Unfortu- nately, details of the types, frequencies and doses/routes of antimicrobialagentsusedonthefarmsstudiedwasnotavailableto determinethepossiblerelationshipbetweenantibioticadminis- tration and the presence of the multi-AR resistant strains. To addressthisconcern,it wouldbedesirabletocarry outfurther

studies,includingfarmswhere antimicrobialagentshave never beenorareinfrequentlyused.

ThemostfrequentserovarobtainedinthisstudywasRissen, followedbyTyphimuriummonophasic(4,(5),12:i:-).InEurope,the prevalence ofS. Typhimurium monophasic(4,(5),12:i:-)causing foodborneoutbreaksanditspresenceinpigsandporkproducts have been increasing (Hopkins et al., 2010; EFSA, 2011).

Additionally, the two major clones (labelled as Spanish and Europeanclones)circulatinginEuropeshowmulti-ARresistance tofourdifferentantimicrobialfamilies,i.e.ampicillin,streptomycin,sulphonamideandtetracycline(ASSuTfamilyproﬁle)(Garcia etal.,2014).Inthisstudy,thenineS.Typhimuriummonophasic strains exhibited the ASSuT phenotype(Table 2), and ﬁve (i.e.

E3G3V4C17, E3G3V4C19, E3G2V4C25, E3G2V4C63 and E3G6V4C13;Fig.1)wererecoveredjustbeforetheanimalswere slaughtered. Continuoussurveillanceshouldbeimplementedto understand the molecular mechanisms and the environmental forcesdrivingtheemergenceandspreadoftheseclonallines.

Conclusions

Multi-ARresistantS.entericawaspresentonpigfarmsinSpain using preventativeveterinary programmesfor thetreatmentof infectiousdiseases,butadirecteffectbetweentheemergenceof thesemulti-ARstrainsandantimicrobialtreatmentsregularlyused onthesefarmswasnotobserved.ThepresenceofARgenesinS.

enterica from pigs was conﬁrmed as being associated to transferrableplasmids.

Conﬂictofintereststatement

Noneoftheauthorsofthispaperhasanyﬁnancialorpersonal relationship with other people or organisations that could inappropriatelyinﬂuenceorbiasthecontentofthepaper.

Acknowledgments

ThisworkwassupportedbyprojectAGL2011-28836fromthe MinisteriodeEconomiayCompetitividadofSpain,theDeparta- mentodeInnovación,EmpresayEmpleodelGobiernodeNavarra (project reference IIQ14064.RI1) and Fundación Caja Navarra, Spain (project reference VATC 2014-0411 y VATC 2015-70628).

Contract of LMG was supported by the Instituto Nacional de InvestigaciónyTecnología AgrariayAlimentaria(INIA)and the EuropeanSocialFund.ContractofVGwassupportedbyFundación Caja Navarra, Spain (project reference VATC 2015-70628). We acknowledgetheCentresdeRecercadeCatalunyaprogrammefor ﬁnancialsupport,andtoJohnG.WildforreviewofEnglish.

Appendix:Supplementarymaterial

Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttps://doi.org/10.1016/j.tvjl.2018.02.002.

References

Aarestrup,F.M.,2004.Monitoringofantimicrobialresistanceamongfoodanimals:

principlesandlimitations.JournalofVeterinaryMedicine.B.Infectious DiseasesandVeterinaryPublicHealth51,380–388.

Anjum,M.F.,Choudhary,S.,Morrison,V.,Snow,L.C.,Mafura,M.,Slickers,P.,Ehricht, R.,Woodward,M.J.,2011.Identifyingantimicrobialresistancegenesofhuman clinicalrelevancewithinSalmonellaisolatedfromfoodanimalsinGreatBritain.

JournalofAntimicrobialChemotherapy66,550–559.

Barton,B.M.,Harding,G.P.,Zuccarelli,A.J.,1995.Ageneralmethodfordetectingand sizinglargeplasmids.AnalyticalBiochemistry226,235–240.

Barton,M.D.,2014.Impactofantibioticuseintheswineindustry.CurrentOpinion inMicrobiology19,9–15.

4 EMA,2015.http://www.ema.europa.eu/docs/en_GB/document_library/Report/

2015/10/WC500195687.pdf(dateaccessed13December2017).