International Journal of Antimicrobial Agents 31S (2008) S4–S8 Relationship between susceptibility to ant factors in paediatric Escherichi d c, A ences, A ology, A o. 5, 06 zmir, Tu lecular Abstract The in vit ains co urinary trac by mu the followin (sfa); necrotizing ce gene increased th © 2007 Else reserv Keywords: Escherichia coli; Antibiotics; Susceptibility; Virulence factor 1. Introdu Pathoge cause a wi ticaemia, n tract infect pathogenic countries, diture [2]. childhood. with anatom patients wi abnormalit colonisatio associated invasins, a ∗ Correspon imleri Faku¨lte Tel.: +90 312 E-mail ad 0924-8579/$ doi:10.1016/j ction nic Escherichia coli strains have the potential to de variety of infectious diseases, including sep- ewborn meningitis, and intestinal and urinary ions [1]. Urinary tract infection (UTI) caused by E. coli remains a common health problem in many resulting in considerable morbidity and expen- UTI is the most common infectious disease in Although it most commonly occurs in patients ically and functionally abnormal urinary tracts, th defects in urine flow caused by urinary tract ies or medical intervention are more susceptible to n with E. coli [3–5]. These strains carry virulence- genes, which may encode toxins, capsules, dhesins and other virulence factors that enable ding author. Present address: Ankara ¨Universitesi, Sag˘lık Bil- si, 06290 Kec¸io¨ren, Ankara, Turkey. 380 81 72 248; fax: +90 312 357 53 23. dress:
[email protected] (M. Arısoy). them to overcome host defences, to proliferate and to cause tissue damage and disease. Urovirulence factors, namely S fimbriae (sfa), afimbrial adhesin I (afaI), haemolysin (hly), cytotoxic necrotizing factor I (cnfI I) and aerobactin (aer), play important roles in the pathogenicity of E. coli strains by overcoming host defence mechanisms and causing disease, and one virulence factor is associated with pyelonephritis (pyelonephritis-associated pili (pap) or P fimbriae) [6,7]. Escherichia coli urovirulence factors can be analysed by multiplex polymerase chain reaction (PCR) and are useful markers for the detection of uropathogenic E. coli [8,9]. Although no conclusive evidence has been established to determine whether or not the presence of certain viru- lence factors is correlated with differences in sensitivities to antimicrobials, recent bacteriological studies regarding pyelonephritis focus mainly on sensitivity to antibiotics and virulence factors [10,11]. The aim of this study was to determine the virulence factors of E. coli and to evaluate the association between potential virulence factors and susceptibility to antimicrobial agents. – see front matter © 2007 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. .ijantimicag.2007.07.030 Mu¨nevver Arısoy a,b,∗, Abbas Yousefi Ra a Ankara University, Faculty of Health Sci b Ankara University, Institute of Biotechn c Mesa Hospital, Clinical Laboratory, Yas¸am Caddesi N d Izmir Municipality Hospital, ˙I e Ankara University, Faculty of Medicine, Department of Mo ro susceptibilities to several antibiotics of 136 Escherichia coli str t infection were analysed. Escherichia coli strains were analysed g virulence factors: pyelonephritis-associated pili (pap); S fimbriae factor I (cnfI); and aerobactin (aer). It was observed that the virulen e sensitivity of susceptible strains. vier B.V. and the International Society of Chemotherapy. All rights imicrobials and virulence a coli isolates li Akın d, Nejat Akar e nkara, Turkey nkara, Turkey 510 So¨g˘u¨to¨zu¨ Ankara, Turkey rkey Genetics, Cebeci, Ankara, Turkey ntaining virulence factors isolated from children with ltiplex polymerase chain reaction for genes encoding afimbrial adhesin I (afaI); haemolysin (hly); cytotoxic s increased antibiotic resistance of resistant strains and ed. M. Arısoy et al. / International Journal of Antimicrobial Agents 31S (2008) S4–S8 S5 2. Materials and methods 2.1. Bacterial strains This stu March and suffering f Departmen the patients from fema catheters. Urine w agar and bl coli were i by biochem of E. coli w 2.2. Extrac Bacteria method as on nutrien suspended ethylenedia at 5000 rpm let was inc (Fermentas lowing day extraction [14]. 2.3. PCR p The E. c and its com Bank (acce (5′-GGT G ECA2 (5′-G which amp 2.4. PCR a Reactio taining the sample, 10 deoxyribon 10 mM of Taq DNA p primers EC in a therm the followi 7 min, foll 1.5 min, an 72 ◦C for 1 Ten microl 1.5% agaro by ethidium bromide staining under ultraviolet (UV) light. PCR a ultiple repo was p llowi is–ED l2, deo 0 mM q DN ers pa G-3′) A-3′) T-3′), A-3′) C-3′), A-3′) C-3′) G- 3′ -3′), a G-3′) -3′) an T-3′) GAT ATA ificati t 94 ◦ ◦C fo ◦C fo imers aer2, s of 32 bp, re volum agar r UV l Antim e anti mined abora (BBL used ; trim mycin floxac uality Statist atistic tests dy included 136 E. coli strains isolated between September 2003 from urine specimens of patients rom UTI admitted to the Pediatric Nephrology t of Ankara University, Turkey. The age range of was 1–16 years and all of the strains were isolated le children. None of the children had indwelling as inoculated onto eosin methylene blue (EMB) ood agar, and >105 colony-forming units/mL of E. solated. Escherichia coli identification was done ical methods and, for confirmation, the wec gene as amplified by PCR [12,13]. tion of bacterial DNA l DNA was extracted by the phenol–chloroform previously described [14]. Briefly, bacteria grown t or EMB agar plates overnight at 37 ◦C were in Tris–EDTA buffer (10 mM Tris HCI, 1 mM mine tetraacetic acid (EDTA)) and centrifuged for 10 min. Following centrifugation, the pel- ubated with Tris–EDTA buffer and proteinase K , Vilnius, Lithuania) overnight at 55 ◦C. The fol- , DNA was extracted by the phenol–chloroform method and suspended in Tris–EDTA buffer rimers for E. coli oli genome organisation within the ECA operon plete DNA sequence were obtained from Gen- ssion no. AE 000455). In this study, primers ECA1 GT CGG CAA GCT TTA TCT CAG-3′) and GT TAA ATT GGG GCT GCC ACC ACG-3′), lify a 763 bp DNA fragment, were used [12]. mplification of E. coli DNA ns were performed in a volume of 50�L con- following components: 5�L of bacterial DNA × buffer solution (Fermentas), 25 mM of MgCl2, ucleotide triphosphate mixture (containing each of dATP, dTTP, dCTP and dGTP), 2 U of olymerase (Fermentas) and 10 pmol each of the AI and ECA2. Amplifications were performed ocycler (Biometra Ltd., Maidstone, UK) under ng conditions: initial denaturation at 95 ◦C for owed by 35 cycles of denaturation at 95 ◦C for nealing at 65 ◦C for 1.5 min and extension at .5 min, and a final extension for 7 min at 72 ◦C. itres of the reaction mixture was analysed on a se gel by electrophoresis. Bands were visualised 2.5. M ously PCR the fo in Tr MgC ing 1 of Ta prim GGC AAT TCA GAC TTA GGC TCT GCC CCG GAA GAG TAT AAG ACC ampl tion a at 94 at 72 Pr aer1– ment 1177 A on 2% unde 2.6. Th deter and L agar were 10�g tobra cipro as a q 2.7. St exact mplification procedures for virulence genes x PCR was performed according to the previ- rted method of Yamamoto et al. [8]. Multiplex erformed in a total volume of 50�L containing ng components: 5�L of bacterial DNA suspended TA, 10 × buffer solution (Fermentas), 25 mM of xyribonucleotide triphosphate mixture (contain- of each of dATP, dTTP, dCTP and dGTP), 2 U A polymerase (Fermentas), 10 pmol each of the p1 (5′-GAC GGC TGT ACT GCA GGG TGT , pap2 (5′-ATA TCC TTT CTG CAG GGA TGC , pap3 (5′-GCA ACA GCA ACG CTG GTT GCA pap4 (5′-AGA GAG AGC CAC TCT TAT ACG , sfa1 (5′-CTC CGG AGA ACT GGG TGC ATC sfa2 (5′-CGG AGG AGT AAT TAC AAA CCT , afa1 (5′-GCT GGG CAG CAA ACT GAT AAC , afa2 (5′-CAT CAA GCT GTT TGT TCG TCC ), aer1 (5′-TAC CGG ATT GTC ATA TGC AGA er2 (5′-AAT ATC TTC CTC CAG TCC GGA , cnf1 (5′-AAG ATG GAG TTT CCT ATG CAG d cnf2 (5′-CAT TCA GAG TCC TGC CCT CAT and 15 pmol each of the primers hly1 (5′-AAC AAG CAC TGT TCT GGC T-3′) and hly2 (5′- TAA GCG GTC ATT CCC GTC A-3′). PCR on was performed in three steps: initial denatura- C for 3 min, followed by 30 cycles of denaturation r 1 min, annealing at 63 ◦C for 30 s and extension r 3 min, and a final extension for 7 min at 72 ◦C. forpap1–pap2, pap3–pap4, sfa1–sfa2, cnf1–cnf2, afa1–afa2 and hly1–hly2 amplified DNA frag- 8 bp, 336 bp, 410 bp, 498 bp, 602 bp, 750 bp and spectively. e of 10�L of the reaction mixture was analysed ose gel by electrophoresis. Bands were visualised ight after ethidium bromide staining. icrobial susceptibility testing microbial susceptibilities of E. coli isolates were by the disk diffusion method of the Clinical tory Standards Institute [15] on Mueller–Hinton , Sparks, MD). Antimicrobial agents (BBL) at the following concentrations: ampicillin, ethoprim/sulfamethoxazole (TMP/SMX), 25�g; , 10�g; cefalothin, 30�g; ceftriaxone, 30�g; and in, 5�g. Escherichia coli ATCC 25922 was used control strain. ical analysis al analyses were performed using χ2 and Fisher’s [16]. S6 M. Arısoy et al. / International Journal of Antimicrobial Agents 31S (2008) S4–S8 Ta bl e 1 Ef fe ct o fv iru le nc e ge ne so n re sis ta nc e an d se n sit iv ity o f1 36 Es ch er ic hi a co li iso la te st o v ar io us an tib io tic s G en e N o. (% )o fs tr ai ns w ith ge ne A m pi ci lli n TM P/ SM X To br am yc in Ce fa lo th in Ce ftr ia xo ne Ci pr ofl ox ac in N o. (% )R N o. (% )S N o. (% )R N o. (% )S N o. (% )R N o. (% )S N o. (% )R N o. (% )S N o. (% )R N o. (% )S N o. (% )R N o. (% )S N V F 55 (40 ) 35 (64 ) 20 (36 ) 27 (49 ) 28 (51 ) 7 (13 ) 48 (87 ) 12 (22 ) 43 (78 ) 4 (7) 51 (93 ) 7 (13 ) 48 (87 ) a er 30 (22 ) 22 (73 ) 8 (27 ) 22 (73 ) 8 (27 ) 7 (23 ) 23 (77 ) 4 (13 ) 26 (87 ) 1 (3) 29 (97 ) 8 (27 ) 22 (73 ) a faI 5 (4) 0 5 (10 0) 5 (10 0) 0 5 (10 0) 0 2 (40 ) 3 (60 ) 1 (20 ) 4 (80 ) 1 (20 ) 4 (80 ) sfa 2 (1) 2 (10 0) 0 1 (50 ) 1 (50 ) 0 2 (10 0) 2 (10 0) 0 0 2 (10 0) 0 2 (10 0) pa p 6 (4) 4 (67 ) 2 (33 ) 4 (67 ) 2 (33 ) 0 6 (10 0) 1 (17 ) 5 (83 ) 1 (17 ) 5 (83 ) 0 6 (10 0) cn fI 5 (4) 3 (60 ) 2 (40 ) 2 (40 ) 3 (60 ) 0 5 (10 0) 1 (20 ) 4 (80 ) 1 (20 ) 4 (80 ) 0 5 (10 0) a er + a faI 5 (4) 5 (10 0) 0 5 (10 0) 0 0 5 (10 0) 0 5 (10 0) 0 5 (10 0) 0 5 (10 0) a er + cn fI 3 (2) 3 (10 0) 0 1 (33 ) 2 (67 ) 0 3 (10 0) 1 (33 ) 2 (67 ) 1 (33 ) 2 (67 ) 0 3 (10 0) a er + pa p 12 (9) 6 (50 ) 6 (50 ) 6 (50 ) 6 (50 ) 0 12 (10 0) 0 12 (10 0) 0 12 (10 0) 1 (8) 11 (92 ) sfa + pa p 3 (2) 2 (67 ) 1 (33 ) 2 (67 ) 1 (33 ) 3 (10 0) 0 1 (33 ) 2 (67 ) 0 3 (10 0) 0 3 (10 0) pa p + cn fI 3 (2) 2 (67 ) 1 (33 ) 1 (33 ) 2 (67 ) 0 3 (10 0) 0 3 (10 0) 0 3 (10 0) 0 3 (10 0) a er + pa p + a faI 2 (1) 2 (10 0) 0 2 (10 0) 0 0 2 (10 0) 0 2 (10 0) 0 2 (10 0) 0 2 (10 0) a er + pa p + cn fI 3 (2) 2 (67 ) 1 (33 ) 1 (33 ) 2 (67 ) 0 3 (10 0) 1 (33 ) 2 (67 ) 0 3 (10 0) 0 3 (10 0) a er + pa p + sfa 2 (1) 0 2 (10 0) 1 (50 ) 1 (50 ) 0 2 (10 0) 0 2 (10 0) 0 2 (10 0) 0 2 (10 0) R ,r es ist an t; S, se n sit iv e; TM P/ SM X ,t rim et ho pr im /su lfa m et ho xa zo le ;N V F, n o v iru le nc e fa ct or s. 3. Results This study focused on 136 E. coli strains causing UTI in children. The in vitro susceptibilities to various antibi- otics of E. coli strains containing pap, sfa, afaI, hly, cnfI and aer virulence factors analysed by multiplex PCR were studied. Fifty-five (40%) of the isolated E. coli strains did not contain virulence factors, whereas 81 (60%) of them contained one or more virulence factors. In the total popula- tion, the number of isolates containing each virulence factor was as follows: 30 (22%) aer; 5 (4%) afaI; 2 (1%) sfa; 6 (4%) pap; 5 (4%) cnfI; 5 (4%) aer + afaI; 3 (2%) aer + cnf; 12 (9%) aer + pap; 3 (2%) sfa + pap; 3 (2%) pap + cnfI; 2 (1%) aer + pap + afaI; 3 (2%) aer + pap + cnfI; and 2 (1%) aer + pap + sfa (Table 1). When resistance to ampicillin of E. coli strains that did not contain virulence genes was compared with those that con- tained at least one virulence factor, resistance was increased in all isolates that contained virulence genes except those with afaI, cnfI, The differe There w between iso gene comb however re isolates har gene comb was increa and sfa + p aer + afaI 100%. The (Fig. 2). Compar lence facto In strains c tobramycin was observ difference When re factors wa Fig. 1. Effect lence factors. aer + pap and aer + pap + sfa gene combinations. nce was statistically significant (P < 0.05) (Fig. 1). as no difference in resistance to TMP/SMX lates containing sfa, aer + pap and aer + pap + sfa inations and isolates with no virulence factors, sistance was decreased by different amounts in bouring aer + cnfI, pap + cnfI and aer + pap + cnfI inations. Moreover, it was found that resistance sed by various amounts in isolates with aer, pap ap gene combinations, and in isolates with afaI, and aer + pap + afaI the resistance rate reached difference was statistically significant (P < 0.05) ing strains with aer with those with no viru- rs, resistance to tobramycin increased two-fold. ontaining afaI and saf + pap genes resistance to reached 100%. For the other gene combinations it ed that resistance to tobramycin disappeared. The was statistically significant (P < 0.001) (Fig. 3). sistance to cefalothin of isolates with no virulence s compared with strains that contained differ- of virulence genes on sensitivity to ampicillin. NVF, no viru- M. Arısoy et al. / International Journal of Antimicrobial Agents 31S (2008) S4–S8 S7 Fig. 2. Effect of virulence genes on sensitivity to trimethoprim/ sulfamethoxazole. NVF, no virulence factors. Fig. 3. Effect virulence fact ent gene co decreased Resistance ing aer, af strains con the strains peared. The (Fig. 4). Fig. 4. Effect lence factors. Fig. 5. Effect of virulence genes on sensitivity to ceftriaxone. NVF, no virulence factors. When resistance to ceftriaxone of strains with no virulence factors was compared with strains having different gene com- ions, resistance was seen to decrease in strains having er gen s harb other rence hen re facto inatio pap g binat the a strain with diffe W lence comb aer + of virulence genes on sensitivity to tobramycin. NVF, no ors. mbinations, it was observed that resistance was in strains harbouring pap and cnfI gene regions. was increased to different amounts in strains hav- aI, aer + cnfI, sfa + pap and aer + pap + cnfI. In taining sfa the resistance rate reached 100%. In with other gene combinations, resistance disap- difference was statistically significant (P < 0.05) of virulence genes on sensitivity to cefalothin. NVF, no viru- increased b gene comb was statisti 4. Discuss In E. c increase in ampicillin, between 1 study are cating that Fig. 6. Effect virulence fact e. Resistance increased to different amounts in ouring afaI, pap, cnfI and aer + cnf, but in strains gene combinations resistance disappeared. The was statistically significant (P < 0.001) (Fig. 5). sistance to ciprofloxacin of strains with no viru- rs was compared with strains with different gene ns, resistance was decreased in strains with the enes. In strains with aer and afaI resistance was y different amounts, and in the strains with other inations resistance disappeared. The difference cally significant (P < 0.001) (Fig. 6). ion oli strains containing certain gene groups, the resistance was between 3% and 36% for between 18% and 51% for TMP/SMX and 1% and 78% for cefalothin. The results of our supported by a study by Orden et al. indi- E. coli virulence factors could be the reason of virulence genes on sensitivity to ciprofloxacin. NVF, no ors. S8 M. Arısoy et al. / International Journal of Antimicrobial Agents 31S (2008) S4–S8 for resistance to different antibiotics [10]. In their study, Katouli et al. found that 53% of E. coli isolates showed high resistance to ampicillin and TMP/SMX antibiotics [17]. In another study by Vranes et al., E. coli strains harbouring the pap gene showed resistance to amoxicillin [18]. In our study, E. coli strains containing the pap gene were also more resistant to ampicillin compared with those not containing the pap gen Maynar factors and strains freq cline and s In our st increased r whilst for o antibiotic w with ceftri combinatio these antib These r ulence fact sensitive to Orden et al In our st strains havi genes, we f in resistant strains. It can toms of ac empirically infection is E. coli. Acknowled Dr Mu¨g with proofr Funding Turkey. Compet Ethical Turkey (No Reference [1] Lalioui a pathog bovine pathogenic Escherichia coli isolates. Infect Immun 2001;69: 937–48. [2] Usein CR, Damian M, Tatu-Chitoiu D, et al. Prevalence of virulence genes in Escherichia coli strains isolated from Romanian adult urinary tract infection cases. J Cell Mol Med 2001;5:303–10. [3] Le Bouguenec C, Archambaud M, Labigne A. Rapid and specific detec- tion of thepap,afaI, and sfa adhesin-encoding operons in uropathogenic Escherichia coli strains by polymerase chain reaction. J Clin Microbiol 1992;30:1189–93. [4] Johnson JR. Virulence factors inEscherichia coliurinary tract infection. lin Mic omberg de´n C. nd recu ithout onnenb scheric nfection ngton, D rskov I yg 197 amamo etectio erase c ohnson N. Char lonal gr lass I an rden JA itro sus airy ca ublic H angon B eneral c n Frenc ayerde or the s eriaceae 002;48: aum E cterizat ntimicr ambroo anual. ory Pres linical or antim ent. W owson T: App atouli . Virul ystitis i ranes J ance to trains o aynard lence a scheric 004;42: e. d et al. studied the relationships between virulence antimicrobial resistance and found that E. coli uently showed resistance to ampicillin, tetracy- ulfonamides [19]. udy, the gene combinations aer, afaI and sfa + pap esistance to tobramycin by between 10% and 87%, ther gene combinations 100% sensitivity to this as noted. A similar situation was also observed axone and ciprofloxacin, where with most gene ns nearly 100% sensitivity was observed against iotics. esults show that E. coli strains with certain vir- ors compared with those not having them were the same antibiotic groups, as in the study of . [10]. udy, when we evaluated the antibiograms of E. coli ng virulence genes and those not having virulence ound that virulence genes increased the resistance strains and increased sensitivity in susceptible be concluded that in patients with symp- ute UTI, antibiotic agents should not be used because there is a significant possibility that caused by a virulent antibiotic-resistant strain of gments e Taner and Mr Samad Joshani-Shirvan helped eading. : Biotechnology Institute of Ankara University, ing interests: None declared. approval: Medical Faculty of Ankara University, . B. 30.2.ANK. 0.01.00.00, 16.01.03*1484). s L, Bouguenec Le C. afa-8 Gene cluster is carried by enicity island inserted into the tRNAPhe of human and C [5] L E a w [6] D E i i [7] O H [8] Y D m [9] J J c c [10] O v d P [11] P g i [12] B f t 2 [13] B a A [14] S m t [15] C f m [16] D C [17] K R c [18] V t s [19] M u E 2 robiol Rev 1991;4:80–128. H, Hellstro¨m M, Jodal U, Leffler H, Lincoln K, Svanborg Virulence-associated traits in Escherichia coli causing first rrent episodes of urinary tract infection in children with or vesicoureteral reflux. J Infect Dis 1984;150:561–9. erg MS, Welch RA. Virulence determinants of uropathogenic hia coli. In: Mobley HLT, Warren JW, editors. Urinary tract s: molecular pathogenesis and clinical management. Wash- C: American Society for Microbiology; 1996. p. 135–74. , Orskov F. Escherichia coli in extra-intestinal infections. J 8;95:551–75. to S, Terai A, Yuri K, Kurazono H, Takeda Y, Yoshida O. n of urovirulence factors in Escherichia coli by multiplex poly- hain reaction. FEMS Immunol Med Microbiol 1995;12:85–90. JR, Stapleton AE, Russo TA, Scheutz F, Brown JJ, Maslow acteristics and prevalence within serogroup O4 of a J96-like oup of uropathogenic Escherichia coli O4:H5 containing the d class III alleles of papG. Infect Immun 1997;65:2153–9. , Ruiz-Santa-Quiteria JA, Garcı´a S, Cid D, De La Fuente R. In ceptibility of Escherichia coli strains isolated from diarrhoeic lves to 15 antimicrobial agents. J Vet Med B Infect Dis Vet ealth 2000;47:329–35. , Chaplain C. Acute pyelonephritis: bacteriological data and ourse of germ resistance. Pathol Biol (Paris) 2003;51:503–7, h. lle P, Zafarullah M. Development of oligonucleotide primers pecific PCR-based detection of the most frequent Enterobac- species DNA using wec gene templates. Can J Microbiol 113–22. Z, Montenegro DA, Licata L, et al. Identification and char- ion of new inhibitors of the Escherichia coli MurA enzyme. ob Agents Chemother 2001;45:3182–8. k J, Maniatis T, Fritsch E. Molecular cloning: a laboratory 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Labora- s; 1989. and Laboratory Standards Institute. Performance standards icrobial susceptibility testing. Eleventh informational supple- ayne, PA: CLSI; 2001. -Sounders B, Tropp RG. Basic clinical biostatistics. Stamford, leton-Lange; 1990, 148–52. M, Brauner A, Haghighi LK, Kaijser B, Muratov V, Mo¨llby ence characteristics of Escherichia coli strains causing acute n young adults in Iran. J Infect 2005;50:312–21. , Scho¨nwald S, Zagar Z. Relation between P-fimbriae and resis- amoxicillin, carbenicillin and tetracycline in uropathogenic f Escherichia coli. Lijec Vjesn 1994;116:178–81, in Croatian. C, Bekal S, Sanschagrin F, et al. Heterogeneity among vir- nd antimicrobial resistance gene profiles of extraintestinal hia coli isolates of animal and human origin. J Clin Microbiol 5444–52. Relationship between susceptibility to antimicrobials and virulence factors in paediatric Escherichia coli isolates Introduction Materials and methods Bacterial strains Extraction of bacterial DNA PCR primers for E. coli PCR amplification of E. coli DNA PCR amplification procedures for virulence genes Antimicrobial susceptibility testing Statistical analysis Results Discussion Acknowledgments References