Predictive analysis of transmissible quinolone resistance indicates Stenotrophomonas maltophilia as a potential source of a novel family of Qnr determinants

BioMed Cen al Page 1 o 14 (page numbe no o ci a ion pu poses) BMC Mic obiology Open Access Resea ch a icle P edic i e analysis o ansmissible quinolone esis ance indica es S eno ophomonas mal ophilia as a po en ial sou ce o a no el amily o Qn de e minan s Ma ía B Sánchez†1, Al a o He nández†1, José M Rod íguez-Ma ínez2, Luis Ma ínez-Ma ínez3,4 and José L Ma ínez*1 Add ess: 1Depa amen o de Bio ecnología Mic obiana, Cen o Nacional de Bio ecnología, CSIC, Da win 3, Can oblanco, 28049-Mad id, and CIBERESP. Spain, 2Depa amen o de Mic obiología, Uni e sidad de Se illa, A . Sanchez Pizjuan SN 41009-Se illa, Spain, 3Se ice o Mic obiology, Uni e si y Hospi al Ma qués de Valdecilla, San ande , Spain and 4Depa men o Molecula Biology, Uni e si y o Can ab ia, San ande , Spain Email: Ma ía B Sánchez - [email p o ec ed]; Al a o He nández - ahd[email p o ec ed]sic.es; José M Rod íguez-Ma ínez - jm od ig[email p o ec ed]; Luis Ma ínez-Ma ínez - u10315@hum .es; José L Ma ínez* - jlm[email p o ec ed] * Co esponding au ho †Equal con ibu o s Abs ac Backg ound: P edic ing an ibio ic esis ance be o e i eme ges a clinical se ings cons i u es a no el app oach o p e en ing and igh ing esis ance o bac e ial pa hogens. To analyse he possibili y ha no el plasmid-encoded quinolone esis ance de e minan s (Qn ) can eme ge and dissemina e among bac e ial pa hogens, we sea ched he p esence o hose elemen s in nea ly 1000 bac e ial genomes and me agenomes. Resul s: We ha e ound a numbe o no el po en ial qn genes in he ch omosomes o aqua ic bac e ia and in me agenomes om ma ine o ganisms. Func ional s udies o he S eno ophomonas mal ophilia Smqn gene show ha plasmid-encoded SmQn con e s quinolone esis ance upon i s exp ession in a he e ologous hos . Conclusion: Al oge he , he da a p esen ed in ou wo k suppo he no ion ha p edic i e s udies on an ibio ic esis ance a e easible, using cu en ly a ailable in o ma ion on bac e ial genomes and wi h he aid o bioin o ma ic and unc ional ools. Ou esul s con i m ha aqua ic bac e ia can be he o igin o plasmid-encoded Qn , and highligh he po en ial ole o S. mal ophilia as a sou ce o no el Qn de e minan s. Backg ound Quinolones a e syn he ic an ibio ics; he e o e, i was hough ha he exis ence o ans e able quinolone esis ance genes in na u e would be unlikely. Resis ance mechanisms o hese d ugs we e expec ed o be only ch omosomally encoded. I was belie ed ha only mu a- ions in he quinolone a ge s (DNA gy ase o opoi- some ase IV) o mu a ions ha led o a pe meabili y dec ease and/o o e p oduc ion o mul id ug esis ance (MDR) e lux pumps would esul in esis ance o hese an ibio ics [1,2]. I was hus assumed ha esis ance o quinolones could no sp ead as he consequence o ho izon al gene ans e Published: 16 Sep embe 2008 BMC Mic obiology 2008, 8:148 doi:10.1186/1471-2180-8-148 Recei ed: 12 Ma ch 2008 Accep ed: 16 Sep embe 2008 This a icle is a ailable om: h p://www.biomedcen al.com/1471-2180/8/148 © 2008 Sánchez e al; licensee BioMed Cen al L d. This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/2.0), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. BMC Mic obiology 2008, 8:148 h p://www.biomedcen al.com/1471-2180/8/148 Page 2 o 14 (page numbe no o ci a ion pu poses) (HGT). La e on, he acquisi ion o quinolone esis ance due o HGT was p oposed as a possibili y based on in i o expe imen s [3]. This was la e con i med in 1998, wi h he desc ip ion o a plasmid-encoded quinolone esis - ance de e minan , ha was named Qn , in clinical isola es [4]. Mo e ecen ly wo o he ans e able quinolone esis - ance de e minan s ha e been desc ibed, he bi unc ional aminoglycoside/quinolones inac i a ing aminoglycoside ace yl ans e ase AAC(6')-Ib-c [5,6] and he quinolone e lux de e minan , QepA [7,8]. Since he disco e y o Qn , he p esence o di e en qn amilies (A, B, S) in esis ance plasmids [9] has been ound wo ldwide in di e en bac e ial pa hogens [10-15]. Mo e ecen ly, plasmid-encoded Qn de e minan s ha e also been desc ibed in en i onmen al isola es o Ae om- onas spp. [16]. The p esence o Qn genes in ch omo- somes has also been shown, including Qn A in Shewanella algae, which is likely he o igin o plasmid-encoded Qn A de e minan s [17], and di e en membe s o he Qn am- ily om Vib ionaceae species [18,19]. The Qn p o eins belong o he pen apep ide epea p o- ein (PRP) amily, which is de ined by he p esence o ep- e i ions in andem o he pa e n (A/C/S/T/V)(D/N)(L/ F)(S/T/R)(G/R) [20-22,17]. A cha ac e is ic ea u e o he Qn p o eins is ha hey a e o med by wo domains o pen apep ide epea s sepa a ed by a single glycine. This s uc u e ma ches a mo i o unknown unc ion named Clus e o O hologous G oup o P o eins (COG) 1357 h p://www.ncbi.nlm.nih.go /COG/. E en hough mem- be s o he PRP amily ha e been iden i ied bo h in p oka yo es and euka yo es [22], qn genes p esen ing he abo e desc ibed COG1357 mo i ha e a much na owe dis ibu ion, mainly in plasmids and in he ch omosomes o some bac e ial species. The amino acid iden i y among di e en Qn p o eins amilies a es be ween 39–60% [23]. Some s udies ha e demons a ed ha Qn p o ec s bo h gy ase and opoisome ase IV om he ac i i y o quinolo- nes [15,24-26]. O he membe s o he PRP amily, ha may play a simila ole include McbG, which p o ec s mic ocin B17-p oducing bac e ia om he ac i i y o his DNA eplica ion inhibi o [27] and M pA [28], a p o ein ha mos likely binds Gy A [22] and p o ides low-le el quinolone esis ance o Mycobac e ium ube culosis. Al hough qn -like elemen s ha con ibu e o in insic quinolone esis ance ha e been desc ibed in he ch omo- some o En e ococcus aecalis [24] and mo e ecen ly in o he G am-posi i e bac e ia [29], hei homology wi h qn genes om G am-nega i e bac e ia is low (a ound 25%), so ha his amily o esis ance elemen s has no been included in ou analysis. I is impo an o no e ha plasmid-encoded quinolone esis ance is mo e p e alen han expec ed conside ing he da e o he i s isola ion, which sugges s ha hese plas- mids ha e been ci cula ing o some ime be o e hey we e i s desc ibed [30]. This highligh s he need o implemen me hods o p edic esis ance be o e i is ecognized wi hin clinical se ings. A me hodology o p edic ing he possibili y o eme gence o a new mechanism o esis ance be o e i appea s in bac- e ial pa hogens has been ecen ly p oposed [31]. The a ailabili y o sequenced genomes allows his ype o analysis. This in o ma ion was used o sea ch o he p es- ence o Qn de e minan s in a ailable sequenced bac e ial genomes and me agenomes. A pu a i e qn gene p esen in he ch omosome o he oppo unis ic pa hogen S eno- ophomonas mal ophilia was chosen o pe o m unc ional analyses. S. mal ophilia is a nosocomial pa hogen in insi- cally esis an o se e al an imic obials [32] due o he ac i i y o an ibio ic-inac i a ing enzymes [33-36] and MDR e lux pumps [37-43]. The possibili y ha ch omosomally-encoded S. mal- ophilia qn gene (he ea e named as Smqn ) could be unc ional in a he e ologous hos has been explo ed. As he esul o ou wo k, we ha e desc ibed new qn genes in he ch omosomes o sequenced bac e ial genomes and ha e ound ha he Smqn gene om S. mal ophilia ende s a low-le el quinolone esis ance pheno ype upon i s exp ession in Esche ichia coli. This s udy demons a es he easibili y o u ilizing cu en ly a ailable da abases along wi h bioin o ma ics and unc ional ools as an app oach o p edic ing esis ance be o e i eme ges in human pa hogens. Me hods Bioin o ma ic ools and nomencla u e o p edic ed Qn p o eins Mul iple p o ein sequence alignmen s we e ca ied ou wi h he p og am Clus alW2 h p://www.ebi.ac.uk/Tools/ clus alw2/index.h ml[56]. Simila i y sea ch was pe - o med wi h he p og am BLAST (NCBI) h p:// www.ncbi.nlm.nih.go /blas /Blas .cgi[57] and genomic- Blas h p://www.ncbi.nlm.nih.go /su ils/ genom_ able.cgi. Iden i ica ion o conse ed mo i s [58] ([59] was done by using Conse ed Domain Da abase and Sea ch Se ice a h p://www.ncbi.nlm.nih.go /S uc u e/cdd/cdd.sh ml. The Qn p o eins alignmen s and homology ees we e ob ained using CLUSTALW2 and Jal iew alignmen edi o wi h de aul pa ame e s [60]. Homology ee was calcula ed using he me hod "A e age dis ance using pe cen age o iden i y" h p:// www.ebi.ac.uk/Tools/clus alw2/index.h ml. The p edic ed Qn p o eins we e named ollowing he ules p oposed in [61]. BMC Mic obiology 2008, 8:148 h p://www.biomedcen al.com/1471-2180/8/148 Page 3 o 14 (page numbe no o ci a ion pu poses) Bac e ial s ains and g ow h condi ions The bac e ial s ains and plasmids used a e shown in Table 1. The s ains we e g own in Lu ia-Be ani (LB) b o h [62] a 37°C unless o he wise speci ied. Suscep ibili y an ibio ic assays The suscep ibili y assays o he di e en an ibio ics we e pe o med in Muelle Hin on b o h (P onadisa) plus Iso- p opyl- hio-ß-D-galac opy anoside 0.5 mM, using he wo old dilu ion me hod in 96-well mic o i e pla es. The esul s we e eco ded a e 48 h o incuba ion a 37°C. To ensu e ha he obse ed changes we e consis en , all Min- imal Inhibi o y Concen a ions (MICs) we e de e mined in h ee independen assays, using di e en bac e ial cul- u es on di e en days. In all cases, a con ol s ain con- aining he plasmid pGEM-T wi hou any inse was included in MICs de e mina ions. In mos cases, he e we e no in e -assay di e ences in he MIC alues. In a ew cases, he e we e one-dilu ion di e ences. Fo he la e , he assay was epea ed one mo e ime o u he assu e assay eliabili y. The quinolones used we e cip o loxacin, enoxacin, ga enoxacin, g epa loxacin, le o loxacin, moxi loxacin, nalidixic acid, no loxacin, o a loxacin and spa loxacin. DNA manipula ions The genomic DNA was ex ac ed using he GNOME® DNA Ki (Q-BIOgene). The PCR Mas e Mix (P omega) was used o ampli y ull-leng h Smqn genes om he di e en S. mal ophilia s ains wi hou hei p omo e sequences. The eac ion con ained 100 ng o genomic DNA o each S. mal ophilia isola e as empla e, and 1 μM o wo di e en se s o speci ic Smqn -p ime s, Qn M+ (5'-CTTGGCAT- GGAATCCCTGAT-3')/Qn M-(5'-TGATGCCTACGGCAC- CAC-3') and Qn MR55+ (5'-CATGGCATGGAATCCCCGA T-3')/Qn MR55-(5'-TGATGTCTACGGCACCAC-3'). We used wo se s o p ime s because he egions a ound qn a e sligh ly di e en in he sequenced S. mal ophilia s ains K279a and R551-3. The eac ions had one dena u a ion s ep a 94°C o 5 minu es, ollowed by 35 ampli ica ion cycles: 94°C o 30 seconds, 61°C o 45 seconds o annealing, and 72°C o 1 minu e o elonga ion, wi h a inal ex ension s ep o 72°C o 5 minu es. The PCR p od- uc s (660 bp) ob ained om he di e en S. mal ophilia s ains (Table 1) we e elec opho esed in 1% aga ose gels wi h TAE, pu i ied om he gel wi h GFX™ PCR DNA and Gel Band Pu i ica ion Ki (GE Heal hca e) and cloned in pGEM-T plasmid (P omega) gene a ing he co espond- ing ecombinan plasmids (Table 1). The ans o ma ion o E. coli KZM120 was made as desc ibed [63]. We chose his pa icula E. coli s ain because i lacks he majo E. coli MDR pump and i has been shown p e iously ha he u iliza ion o MDR-de ec i e s ains acili a es he cha ac- e iza ion o low-le el mechanisms o esis ance [39]. The ans o man s we e selec ed in LB aga wi h 100 μg/ml ca benicillin. The ecombinan plasmids we e pu i ied using Wiza d® Plus SV Minip ep Ki (P omega) and he inse o each plasmid was sequenced using he uni e sal p ime s M13 o wa d and M13 e e se by Secugen S.L. h p://www.secugen.es/ o con i m he iden i y o he sequence and es ablish he o ien a ion o he qn gene. The sequences o he di e en Smqn genes ha e been deposi ed a GenBank wi h numbe s om EU681371 o EU681385. Only hose plasmids con aining he Smqn gene in he igh o ien a ion o allow exp ession om he pGEM-T lac p omo e we e used in he unc ional assays. Iden i ica ion o Smqn in di e en S. mal ophilia isola es The po en ial p esence o Smqn in bo h clinical and en i- onmen al S. mal ophilia s ains was e alua ed by PCR using he p ime s qn I1 (5'-AGAAAGTGGTCGACCAG- CAG-3')/qn I2 (5'-GCAGGTTCGACTTCTTGATG-3') and qn I3 (5'-CAACGCCAGCTTCATGAACC-3')/qn I4 (5'- AGTTGGCGCTGTTCCAGTCG-3'), which ampli y 312 bp and 220 bp agmen s o wo in e nal egions in he Smqn gene o S. mal ophilia espec i ely. The PCR was made as desc ibed abo e, wi h he ollowing p og am, one dena u a ion cycle a 94°C o 5 minu es, ollowed by 35 ampli ica ion cycles: 94°C o 30 seconds, 55°C o 45 seconds o annealing, and 72°C o 30 sec- onds o polyme iza ion, wi h a inal ex ension o 72°C o 5 minu es. The PCR p oduc s we e analyzed in 1% aga- ose gels wi h TAE. Exp ession and iden i ica ion o he SmQn p o ein The amoun o SmQn p o ein was es ima ed by SDS- PAGE [63] and Coomassie blue s aining, using cell ex ac s om bac e ia g own in LB a 37°C o e nigh . The amoun o p o eins loaded in each lane was no malized o a ound 4 × 107 cells. In all cases, he global amoun o p o eins was equal in all lanes o a gi en gel as can be seen a e s aining. This se es as a supplemen a y loading con- ol [64]. To ensu e ha he p o ein wi h he p edic ed molecula size was indeed SmQn , he co esponding band was excised and iden i ied, as desc ibed [65,39] by Pep ide Mass Finge p in ing using MS-MALDI TOF (P o eo ed: h p://www.p o eo ed.o g/). E hical conside a ions This in es iga ion did no equi e e hical clea ance. Resul s and Discussion The iden i ica ion in en i onmen al mic oo ganisms o pu a i e an ibio ic esis ance elemen s [44] ha migh ans e in he nea u u e o pa hogenic bac e ia is an impo an opic ha we a e jus beginning o add ess [31]. BMC Mic obiology 2008, 8:148 h p://www.biomedcen al.com/1471-2180/8/148 Page 4 o 14 (page numbe no o ci a ion pu poses) To ha goal a combina ion o bioin o ma ic and unc- ional ools can gi e in o ma ion o p edic esis ance be o e i eme ges. Following his concep , we ha e explo ed he p esence o pu a i e qn genes in a ailable da abases o bac e ial genome sequences, because he gene ic con ex o plasmidic qn genes suppo s he idea ha hey ha e been acqui ed by HGT. Fo ins ance, he plasmid-encoded qn A gene has been ound in a sul1- ype in eg on nea an ISCR1 elemen [21,45,25]. A simila s uc u e was ound in qn B2 [46], and qn S is adjacen o a Tn3 ansposon s uc u e [10,11]. In con as , i has been epo ed ha ch omosomal qn A, in Shewanella algae [17] and se e al qn -like genes iden i ied wi hin he Vib i- onacea amily [18] a e no linked o hese mobile gene ic elemen s. This sugges s ha ch omosomally-encoded qn genes ha e no been acqui ed ecen ly by HGT because o ecen an ibio ic selec i e p essu e and ha ye uniden i ied ch omosomally-encoded qn genes could be a sou ce o new ans e able quinolone esis ance genes. Bioin o ma ic sea ch o pu a i e qn genes To add ess he p esence o pu a i e qn genes in he ch o- mosomes o bac e ia ha could se e as ese oi s o his amily o quinolone esis ance de e minan s, a bioin o - ma ic i e a i e sea ch by sequence homology o se e al alleles o Qn A, Qn B and Qn S was conduc ed agains genomic o Whole Genome Sho gun Sequence (WGS) da abases a he NCBI home page h p:// www.ncbi.nlm.nih.go /. Mos o he pu a i e qn genes ound in his sea ch ha e been p e iously anno a ed as hypo he ical p o eins con aining he COG1357 mo i . As shown in Figu e 1, ch omosomally-encoded Qn p o eins p esen a la ge deg ee o homology. Pu a i e qn genes we e ound in 22 ou o he 960 genomes es ed. The species con aining pu a i e qn genes belong o eigh di e en gene a (Figu e 2). Mos o hem inhabi an aqua ic en i onmen (in some ins ances he deep sea) sugges ing ha plasmidic qn may ha e been o igina ed om aqua ic bac e ia [17,19,18]. Since he la ge majo i y o bac e ial species ha e no been cul u ed so a , me agenomic s udies allow a mo e powe - ul sea ch o ele an genes in di e en en i onmen s by means o non-cul u e based me hodologies. Thus, besides analysing sequenced genomes, a ailable me agenomic sequence da abases h p://www.ebi.ac.uk/ as a33/ wgs.h ml we e also sea ched o he p esence o qn genes. Ou s udy showed ha he me agenomes om sea-wa e [47] con ain genes homologous o qn (he ea e named as M gqn ), u he suppo ing he no ion ha qn genes a e o igina ed om aqua ic mic oo ganisms. One o he qn sequences ound in his me agenomic sea ch was highly homologous o plasmid-encoded Qn B [15] p o- eins (Figu e 3). I has been shown ha S. algae is likely he o igin o plasmid-encoded qn A genes [17]. Ou anal- yses, s ongly sugges ha plasmid-encoded qn B has also o igina ed om a cu en ly unknown, ma ine mic oo - ganism. The egions su ounding he ch omosomal qn genes ound in ou bioin o ma ics analysis we e analysed o asce ain whe he , based in he le el o sin eny, he o igin o hese qn de e minan s is likely monophyle ic o polyphyle ic. As shown in Figu e 2, and in Addi ional ile 1, he egions a e e y simila o some o he analysed species belonging o he same genus, al hough clea di e - ences a e also obse ed o o he species. Fo ins ance, he s uc u e su ounding qn in Vib io is e y simila o h ee species, whe eas he s uc u e in i e o he species and in Pho obac e ium p o undum is di e en . The high le el o obse ed sin eny in hese egions, sug- ges s ha qn was acqui ed by Vib io pa ahaemoly icus, Vib io alginoly icus and Vib io ha eyi be o e hei di e - gence, and was acqui ed by he o he Vib ionaceae a e hei di e gence. This idea i s well wi h he ela ionship o Vib io Qn p o eins (Figu e 4). Sequence da a we e inspec ed o he p esence o elemen s po en ially in ol ed in he ans e o qn genes. As shown in Figu e 2, pu a i e ansposases o in eg ases we e ound in he egions su ounding he qn de e minan s in some Shewanella species, and one pu a i e ecombinase was ound as well in Mo i ella sp. Whe he hose elemen s a e emnan o o me gene ans e e en s o cons i u e a isk o he ans e ence o hose qn genes h ough HGT o a new hos , emains o be es ablished. In any case, he analyses o a ailable sequences o bac e- ial genomes sugges ha qn genes a e ancien elemen s in he bac e ial ch omosomes o a speci ic subse o bac e- ial species. No ewo hy, hose genes a e equen ly lanked by genes coding o pu a i e e lux pumps (Figu e 2), a si ua ion ha sugges s ha Qn may ha e unc ional oles in de oxi ica ion p ocesses in wa e -dwelling bac e- ia. Al hough his linkage has no been so a obse ed o plasmid-encoded qn genes, he coope a ion o wo di e - en mechanisms o esis ance migh be an impo an ele- men in cu en esis ance o quinolones. A pu a i e qn gene is p esen in he ch omosome o S. mal ophilia Once he bioin o ma ic analysis was made, we ocused ou e o s in he s udy o he o ganism wi h highes clin- ical ele ance. Among he o ganisms ca ying pu a i e BMC Mic obiology 2008, 8:148 h p://www.biomedcen al.com/1471-2180/8/148 Page 5 o 14 (page numbe no o ci a ion pu poses) Homology o p edic ed new ch omosomally-encoded Qn p o einsFigu e 1 Homology o p edic ed new ch omosomally-encoded Qn p o eins. The Figu e shows he alignmen o ch omosoma- lly-encoded pu a i e Qn p o eins om he analysis o a ound 1000 genomes and me agenomes. The igu e includes one mem- be o each plasmidic Qn p o ein amily (highligh ed in ed), so ha he simila i ies a e clea e . The pu a i e Qn p o ein om he sea me agenome is also included (M gQn ). As shown in he Figu e, conse a ion is high along he p o eins, al hough he simila i ies a e lowe a hei N- e minus. The au oma ed anno a ion o Swoqn p esen s wo s op codons. Since we igno e whe he his is a sequence mis ake o Swoqn is a pseudogene, a manual ansla ion o i s nucleo ide sequence was pe o med o his alignmen . S op codons a e ep esen ed wi h an X. Accession numbe s o he p o eins used o he alignmen a e: AhQn (YP_854820.1), AsQn (YP_001143795.1), MsQn (ZP_01897167.1), Qn A1 (ACA43024), Qn B1 (ABG82188), Qn S1 (ABU86826), SbeQn (ZP_02157732.1), S Qn (YP_750786.1), SliQn (ABO22341.1), SmQn (ZP_01643096.1), SpaQn (ABV89003.1), Sp Qn (YP_001478290.1), SseQn (ABV35251.1), SwoQn (ZP_01539189.1), PpQn (CAG21998.1), PsQn (EAS39797.1), ValQn (ZP_01261394.1), VanQn (ZP_01234687.1), V Qn (ZP_02136549.1), VhQn (ZP_01985396.1), VpQn (ZP_01990287.1), VshQn (ZP_01866232.1), VspQn (ZP_00989608.1), V Qn (AAO07889.1) and M gQn (AACY020347520). BMC Mic obiology 2008, 8:148 h p://www.biomedcen al.com/1471-2180/8/148 Page 6 o 14 (page numbe no o ci a ion pu poses) P edic ed qn genes in genome da abasesFigu e 2 P edic ed qn genes in genome da abases. The s uc u e o he egions su ounding p edic ed qn genes (in ed) is shown. Elemen s ha migh con ibu e o ans e o hose genes a e highligh ed in black. The p esence o pu a i e MDR sys- ems is also highligh ed. WGS (whole genome sho gun), CG (Comple e genome), Ch (Ch omosome). Only hose pu a i e qn de e minan s showing mo e han 45% iden i y wi h known plasmid-encoded Qn A o Qn B de e minan s we e included in he analysis. The abb e ia ions used in he Figu e a e shown in Addi ional ile 1. BMC Mic obiology 2008, 8:148 h p://www.biomedcen al.com/1471-2180/8/148 Page 7 o 14 (page numbe no o ci a ion pu poses) uncha ac e ized qn genes we chose S. mal ophilia, which is an in insically esis an oppo unis ic pa hogen esponsible o 4.3% o G am-nega i e in ec ions among in ensi e ca e uni s pa ien s in he USA [48]. No ewo hy, in S. mal ophilia esis ance o luo oquinolo- nes (pa icula ly o less ac i e ones, such as no loxacin) is no always accompanied by high le el esis ance o nalid- ixic acid, which is he usual and classical associa ion in En e obac e ia. In e es ingly, esis ance o luo oquinolo- nes and suscep ibili y o nalidixic acid in S. mal ophilia esemble he pheno ype ecen ly obse ed by Cano ME e al (unpublished esul s) in some En e obac e iaceae con- aining qn genes. A pu a i e qn gene (Smqn ) was iden i- ied in he wo a ailable ull-genome sequences o S. mal ophilia: he clinical isola e K279a h p:// www.sange .ac.uk/P ojec s/S_mal ophilia/[49], and he en i onmen al s ain R551-3, h p://genome.jgi-ps .o g/ d a _mic obes/s ema/s ema.home.h ml. As shown in Figu e 2, he genomic s uc u e a ound Smqn was he same in bo h s ains, in spi e hei di e en (clinical and en i onmen al) sou ce o isola ion. This high deg ee o sin eny, oge he wi h he lack o ele- men s in ol ed in HGT nea Smqn sugges s ha his gene has an ancien o igin in S. mal ophilia. Fu he con i ma- ion o his likely monophyle ic o igin has been ob ained om he analyses o he Smqn sequence om di e en S. mal ophilia isola es in his s udy. The p edic ed ansla ion o Smqn sequence is a 219 amino acid p o ein (Figu e 5), con aining andem epe i ions o he pa e n (A/C/S/T/ V)(D/N)(L/F)(S/T/R)(G/R) ypical o he pen apep ide amily [20,18,22], wi h wo domains o pen apep ide epea s sepa a ed by a single glycine, ha co esponds o a COG1357 mo i . P e alence o he Smqn gene in S eno ophomonas mal ophilia s ains Since Smqn migh be in ol ed in quinolone esis ance, we wan ed o asce ain he p e alence o his gene in S. mal ophilia popula ions. To ha goal, and since he sequences o he qn genes o he s ains K279a and R551- 3 we e sligh ly di e en , wo pai s o oligonucleo ides Qn M+/- and Qn MR55+/- we e designed o ampli y Smqn genes om di e en S. mal ophilia s ains. The se o analysed s ains comp ised bo h clinical and en i on- men al S. mal ophilia isola es (Table 1). The SmQn sequences om he di e en s ains display a high amino acid iden i y, anking om 94 o 100% (Fig- u e 5). Howe e he amino acid iden i y o he SmQn p o ein encoded in he ch omosome o he sequenced S. mal ophilia s ain K279a, is low compa ed wi h o he known Qn p o eins, sha ing 38%, 59%, 38% and 41% wi h Qn A1 (AY070235), Qn B2 (DQ351241), Qn S2 (AB187515) and V Qn (DQ889870) espec i ely. Fo his eason we conside ha Smqn ep esen s a possible new sub ype o qn genes. Clus al analysis (Figu e 4) indi- ca es ha he SmQn p o ein clus e s wi h plasmidic Qn B p o eins, al hough o ms a sepa a e b anch. The e o e, ou esul s sugges ha SmQn is he i s cha ac e ized ch o- mosomally encoded Qn B-like p o ein. We we e unable o ampli y he ull qn gene om he iso- la es E729, F227, E301, F861 and e-p5. Since qn sequences whe e sligh ly di e en in K279a and R551-3, as well as in he isola es analysed in he p esen s udy, wo new pai s o oligonucleo ides (qn I1/qn I2 and qn I3/ qn I4) we e used o ampli y an in e nal conse ed egion o Smqn . Posi i e ampli ica ion was ob ained in all cases (no shown) indica ing ha all S. mal ophilia isola es con- ained he qn gene. Alignmen o he plasmid-encoded p o ein Qn B wi h a pu a i e Qn p o ein encoded in a ma ine me agenomeFigu e 3 Alignmen o he plasmid-encoded p o ein Qn B wi h a pu a i e Qn p o ein encoded in a ma ine me agen- ome. The pe cen age o iden i y be ween bo h p o eins was 94.9%, wi h a homology o 99.5%. Top Qn B5, bo om Qn p o- ein om a ma ine me agenome (M gQn ). Two do s indica e iden ical amino acids, one do homologous amino acids. BMC Mic obiology 2008, 8:148 h p://www.biomedcen al.com/1471-2180/8/148 Page 8 o 14 (page numbe no o ci a ion pu poses) Func ional analysis o SmQn as a quinolone esis ance de e minan in a he e ologous hos To asce ain whe he SmQn migh con ibu e o qui- nolone esis ance in a he e ologous hos , he di e en alleles o he qn gene ob ained om di e en S. mal- ophilia s ains we e cloned in o he plasmid pGEM-T. Co esponding ecombinan plasmids, con aining he Smqn gene, we e exp essed in E. coli, and he quinolone suscep ibili y o s ains exp essing SmQn was compa ed o ha o he isogenic E. coli s ain. We ound ha in ans exp ession o SmQn om di e en ecombinan plasmid esul s in a 2 o 32- ols dec ease in he quinolone suscep- ibili y o E. coli (Table 2). No ewo hy, di e en ecombinan s ains p esen ed sligh , bu consis en , di e ences in hei suscep ibili y o se e al quinolones (Table 2). I has been p e iously s a ed ha ansconjugan s o plasmids con aining he qn gene p esen dispa a e le els o quinolone esis ance, likely due o di e en le els o exp ession o he Qn p o ein in he Clus al analysis o Qn p o einsFigu e 4 Clus al analysis o Qn p o eins. Cha ac e ized and p edic ed Qn p o eins we e clus e ed. The ee was calcula ed using he a e age pe cen age o iden i y. O he wo membe s o he pen apep ide epea amily wi h p o en esis ance o luo oqui- nolones, M pA om Mycobac e ium ube culosis and McbG om Esche ichia coli, we e included in he analysis, bu a e no shown because hey plo ed ou o he Qn ee. Ch omosomally-encoded Qn p o eins a e highligh ed in bold. The posi ion o he SmQn p o ein om S. mal ophilia as a membe o he Qn B clus e is highligh ed wi h a squa e. No el Qn p o eins a e named acco ding o he species whe e hey a e o igina ed: AhQn (Ae omonas hyd ophila), AsQn (Ae omonas salmonicida), MsQn (Mo i ella sp), Sp Qn (Se a ia p o eamaculans), SbeQn (Shewanella ben hica), S Qn (Shewanella igidima ina), SliQn (Shewanella loihica), SpaQn (Shewanella pealeana), SseQn (Shewanella sediminis), SwoQn (Shewanella woodyi), SmQn (S eno o- phomonas mal ophilia), PsQn (Psych omonas), ValQn (Vib io alginoly icus), VanQn (Vib io angus um), V Qn (Vib io ische i), VhQn (Vib io ha eyi), VshQn (Vib io shilonii), VspQn (Vib io splendidus), V Qn (Vib io ulni icus). BMC Mic obiology 2008, 8:148 h p://www.biomedcen al.com/1471-2180/8/148 Page 9 o 14 (page numbe no o ci a ion pu poses) Alignmen o deduced amino acid sequences o di e en SmQn p o einsFigu e 5 Alignmen o deduced amino acid sequences o di e en SmQn p o eins. The amino acids ha changed in he di - e en SmQn alleles a e highligh ed in a g ey squa e. The glycine esidue (G) sepa a ing he wo domains o pen ape ide epea s, is ma ked wi h a box. di e en ansconjugan s [50]. To asce ain whe he his could be he si ua ion wi h he di e en SmQn -exp ess- ing plasmids, he exp ession o he SmQn p o ein was es ima ed in he di e en E. coli ans o man s. As shown in Figu e 6A, ans o man s con aining ei he pBS3.25 o pBS3.13 exp essed highe le els o SmQn han he o he ans o man s. Du ing ou wo k, we de ec ed ha some Qn -exp essing plasmids we e los a e subcul u ing e en in he p esence o ampicillin, al hough hey could be main ained in he p esence o ca benicillin. Plasmid loss a e he bac e ia has deg aded he selec i e an ibio ic is an indica ion o physiological bu den [51]. Ou esul s sugges ha high- le el exp ession o SmQn is likely ha m ul o E. coli. I is possible ha , as desc ibed o o he sys ems [52], a as adap a ion be ween he hos cell and he plasmid occu ed ha esul s in educed le els o SmQn and ha his allowed some o he clones o o e come he pu a i e SmQn oxici y. To add ess his possibili y, E. coli KZM120 was e ans o med wi h he plasmids pBS3.8, pBS3.25 and pBS3.13. Two di e en clones om each e ans o ma ion we e chosen and hei suscep ibili y o quinolones es ed. As shown in Table 3, an o e all educ- ion in MICs as well as in he le el o SmQn exp ession (Figu e 6B) was obse ed o he new ans o man s con- aining he plasmids pBS3.8, pBS3.25 and pBS3.13. How- e e , he pheno ype o educed suscep ibili y o quinolones was main ained. Inc eased esis ance due o highe exp ession o an ibio ic esis ance genes has been desc ibed mainly associa ed wi h plasmidic be a-lac ama- ses [53-55]. Ou esul s oge he wi h al eady published da a indica e ha gene-dosage migh be ele an o qn - media ed quinolone esis ance. Conclusion By using a combina ion o bioin o ma ics and unc ional ools, we ound a numbe o no el pu a i e qn genes in he ch omosomes o aqua ic bac e ia and in me agen- omes om ma ine o ganisms. This u he suppo s he no ion ha he o igins o qn de e minan s a e in wa e - dwelling bac e ia. I has been ecen ly shown ha Ae om- onas spp. ob ained om wa e samples con ain plasmid- encoded qn genes [16]. These esul s ha e been in e - p e ed as a consequence o he p esence o quinolones in i e s ha d i e he e olu ion o aqua ic bac e ia owa ds quinolone esis ance. This explana ion is likely ue o plasmid-encoded qn genes. Howe e , he p esence o qn -