International Journal of Antimicrobial Agents 22 (2003) 626 /629 www.ischemo.org
Prevalence and characterization of the mechanisms of macrolide, lincosamide and streptogramin resistance in viridans group streptococci Sofia Ioannidou a,b, Joseph Papaparaskevas b, Panayotis T. Tassios b, Maria Foustoukou a, Nicholas J. Legakis b, Alkiviadis C. Vatopoulos c,* a Department of Clinical Microbiology, P. and A. Kyriakou Childrens’ Hospital, Athens, Greece Department of Microbiology, Medical School, National and Kapodistrian University of Athens, Athens, Greece c Department of Hygiene and Epidemiology, Medical School, National and Kapodistrian University of Athens, Mikras Asias 75, 115 27 Athens, Greece b
Received 26 February 2003; accepted 8 May 2003
Abstract The presence of erm genes conferring constitutive and inducible resistance, as well as that of the mefA gene conferring only constitutive resistance, was investigated using PCR in 70 erythromycin resistant (MIC ]/1 mg/l) strains of viridans group streptococci (VGS) (18 Streptococcus mitis biotype 1, 16 S. mitis biotype 2, 15 S. oralis , 12 S. salivarius and nine S. sanguis ) isolated from the oropharynx of healthy Greek children. All of the 56 isolates belonging to resistance phenotype M harbored the mefA gene. All of the 14 isolates constitutively resistant to macrolides and lincosamides (phenotype CR) harbored the ermB gene. Co-presence of both genes was not observed, whereas class A erm gene (previously known as ermTR ) was not detected. Our results are consistent with a possible role of VGS as a reservoir of resistance genes now prevalent in pathogenic species of streptococci. # 2003 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. Keywords: Viridans streptococci; mefA , ermB ; Class A erm gene; Macrolides
1. Introduction During recent years, it has become increasingly apparent that the future of antibiotic treatment will depend on the evolution of resistance not only among the pathogens targeted by therapy, but also among commensal organisms, which could function as reservoirs of resistance genes [1]. The increasing numbers of erythromycin-resistant (EryR) Streptococcus pyogenes and S. pneumoniae constitute a major public health problem in many parts of the world today [2,3]. Viridans Group Streptococci (VGS), comprising up to 60% of the oral bacterial flora, are considered to be a possible reservoir for the dissemination of macrolide resistance genes to the above pathogenic streptococcal species [1,4]. In a previous
* Corresponding author. Tel.: /30-210-746-2071; fax: /30-210746-2079. E-mail address:
[email protected] (A.C. Vatopoulos).
communication, a 38.5% resistance rate to erythromycin in VGS recovered from the oropharynx of healthy Greek children was reported [5]. The aim of the present study was to provide additional information to existing data from other countries, regarding the molecular mechanisms of erythromycin resistance in this group of bacteria in Greek children.
2. Materials and methods From a total of 77 erythromycin resistant (EryR) VGS isolates (MIC /1 mg/l) described in our previous communication [5], 70 were included in the present study (seven isolates were not recovered after storage). Briefly, strains were isolated from oropharyngeal swabs of 96 children, aged 3 months to 14 years old (median age 7 years), attending the orthopaedic outpatient department of the hospital. The children were not suffering from respiratory tract infection and had not
0924-8579/03/$30 # 2003 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/j.ijantimicag.2003.05.001
S. Ioannidou et al. / International Journal of Antimicrobial Agents 22 (2003) 626 /629
received any antibiotic therapy or prophylaxis during the previous 3 months. The group of isolates included 18 Streptococcus mitis biotype 1, 16 S. mitis biotype 2, 15 S. oralis , 12 S. salivarius and nine S. sanguis. Determination of MLSB resistance phenotypes was based on the double disk test using erythromycin and clindamycin, as previously described [6]. Briefly, strains resistant to erythromycin but susceptible to clindamycin were allocated to the M phenotype, whereas those resistant to both antibiotics were allocated to the MLSB phenotype; to the CR MLSB phenotype if they expressed resistance to clindamycin constitutively, or to IR MLSB phenotype if they expressed resistance to clindamycin inducibly. For DNA extraction, a modification of a previously described protocol was used [7]. PCR amplification of segments of the ermB and mefA genes was performed using protocols and conditions described previously [8,9]. Strains VR5 and VR3, containing the ermB and mefA genes, respectively, were used as positive controls. For the PCR amplification of class A erm gene (ermTR ) segment, primers ermTR1 (5?-TTG TTC ATT GGA TAA TTT ATC-3?) and ermTR2 (5?-CTT GTG GAA ATG AGT CAA CGG-3?) were used [10] (J. Jalava, personal communication). A total of 50 ng of template DNA was added to a mixture containing 20 pmol of each primer, 0.5 U of Taq polymerase, 200 mM of dNTPs, 1.5 mM MgCl2 and 1 / PCR buffer, in a final volume of 25 ml. Strain KUOR3 was used as a positive control. Conditions were as follows: denaturation at 94 8C for 4 min, followed by 35 cycles of denaturation at 94 8C for 1 min, annealing at 51 8C for 1 min and elongation at 72 8C for 1 min (J. Jalava, personal communication). All PCR reactions were performed in a PowerBlock thermocycler (Ericomp Inc., San Diego, CA 92120). The products were resolved by electrophoresis on 1% agarose gels, stained with ethidium bromide and recorded under UV illumination. The expected amplicon sizes were 640 bp for ermB , 940 bp for mefA and 552 bp for class A erm gene.
3. Results and discussion PCR amplification of erythromycin resistance gene segments (Table 1), revealed that all of the 56 M phenotype isolates harbored the mefA gene. All 14 CR MLSB phenotype isolates harbored the ermB gene. All genotypic results were in accordance to the resistance phenotypes. The mefA gene was found to be predominant in all species of VGS, except for S. salivarius , where an equal distribution of the two genes was detected. Neither class A erm gene, nor the simultaneous presence of two genes was observed.
627
Although studies of the prevalence of different macrolide resistance genes among VGS, in various parts of the world, have produced diverse results, the mefA gene has also been found to prevail by various authors. MefA and mefE genes are no longer considered to be different and only mefA is described in [11]. Surveys from Spain [4] and the UK [12] indicated that among VGS strains isolated from patients suffering from pharyngitis and respiratory tract infections, the M phenotype and mefA gene presence predominated. However, another study from Spain [13] reported equal numbers of CR MLSB and M phenotypes, with different distributions being detected among the different species: the CR MLSB phenotype predominated in S. oralis and S. sanguis, whereas the M phenotype predominated in S. mitis and S. salivarius . Additionally, among EryR VGS clinical isolates from invasive diseases in Spain, M phenotype and the mefA gene was detected in 48%, while phenotype CR MLSB and the ermB gene was found in 52% of the isolates [14]. Finally, both ermB and mefA genes were found in a limited number of EryR S. mitis and S. oralis clinical isolates from Japan [15]. Nevertheless, in other studies, an increased prevalence of MLSB phenotype was documented. In Portugal, among 29 erythromycin resistant VGS, the ermB gene was detected in nine, the mefA in two, whilst no gene was detected in 18 isolates [8]. Moreover, the ermB gene predominated among 22 EryR S. anginosus (‘S. milleri ’) clinical isolates, whilst no resistance gene was detected in three isolates [16]. In France, among 35 EryR VGS clinical isolates, 23% were of the M phenotype, due to the mefA gene, whilst the remaining 77% were of the CR MLSB phenotype [17]. The difference in the isolation frequencies among the various VGS species, as well as the differences in resistance phenotypes of the respective genes might be related to the variable selective pressure in the respective ecological niches. Interestingly our results were in agreement with other studies in this country, indicating that irrespective of the streptococcal species, M phenotype and the underlying mefA expression constitutes the predominant erythromycin resistance mechanism among different species of streptococci in Greece, found in more than 50% of both S. pyogenes and S. pneumoniae [18 /21]. On the other hand, and with respect to the MLSB phenotype, the CR phenotype and the ermB gene is most frequently detected among the S. pneumoniae MLSB clinical isolates [21] (N.J. Legakis, unpublished results), whereas the IR phenotype and the class A erm gene seems to be more prevalent among the S. pyogenes MLSB clinical isolates [20,21]. In that respect, and since in most cases a carrier state in the oropharynx precedes S. pyogenes and S. pneumoniae infection [22] our finding is consistent with a possible role of this group of normal flora bacteria as a reservoir of the mefA gene now prevalent in patho-
S. Ioannidou et al. / International Journal of Antimicrobial Agents 22 (2003) 626 /629
628
Table 1 MLSB phenotypes and resistance genes of various species of viridans streptococcal isolates Species
S. mitis 1 S. mitis 2 S. oralis S. salivarius S. sanguis Total
Number tested
18 16 15 12 9 70
MLSB phenotype
M phenotype
Number (%)
Number of erm B (/)
Number (%)
Number of mef A (/)
3 (17) 1 (6) 3 (20) 6 (50) 1 (11) 14 (20)
3 1 3 6 1 14
15 (83) 15 (94) 12 (80) 6 (50) 8 (89) 56 (80)
15 15 12 6 8 56
genic genera of streptococci. The absence of the class A erm gene from the VGS isolates tested must be emphasized, and a natural reservoir for this gene must be further investigated, as also suggested by other authors [23].
Acknowledgements We gratefully acknowledge the provision of control strains VR5 and VR3 by G. Cornaglia (Institute of Microbiology, University of Verona, Italy) and for control strain KUOR3 by P. Huovinen (National Public Health Institute, Turku, Finland). We also thank J. Jalava (National Public Health Institute, Turku, Finland) for providing the sequences for primers ermTR1 and ermTR2 and the conditions for the PCR.
References [1] Bryskier A. Viridans group streptococci: a reservoir of resistant bacteria in oral cavities. Clin Microbiol Infect 2002;8:65 /9. [2] Amsden GW. Pneumococcal macrolide resistance */myth or reality. J Antimicrob Chemother 1999;44:1 /6. [3] Efstratiou A. Group A streptococci on the 1990s. J Antimicrob Chemother 2000;45(Suppl.):3 /12. [4] Aracil B, Minabres M, Oteo J, Torres C, Gomez-Garces JL, Alos JI. High prevalence of erythromycin-resistant and clindamycinsusceptible (M phenotype) viridans group streptococci from pharyngeal samples: a reservoir of mef genes in commensal bacteria. J Antimicrob Chemother 2001;48:592 /4. [5] Ioannidou S, Tassios P?, Kotsovili-Tseleni A, Foustoukou M, Legakis NJ, Vatopoulos AC. Antibiotic resistance rates and macrolide resistance phenotypes of viridans group streptococci from the oropharynx of healthy Greek children. Int J Antimicrob Agents 2001;17:195 /201. [6] Wu JJ, Lin KY, Hsueh PR, Liu JW, Pan HI, Sheu SM. High incidence of erythromycin-resistant streptococci in Taiwan. Antimicrob Agents Chemother 1997;41:844 /6. [7] Widjojoatmodjo MN, Fluit AC, Verhoef J. Rapid identification of bacteria by PCR single-strand conformation polymorphism. J Clin Microbiol 1994;32:3002 /7. [8] Luna VA, Coates P, Eady A, Cove EA, Nguyen TT, Roberts MC. A variety of Gram-positive bacteria carry mobile mef genes. J Antimicrob Chemother 1999;44:19 /25.
[9] Roberts MC, Chung WO, Roe D, Xia M, Marguez C, Borthagaray G, et al. Erythromycin-resistant Neisseria gonorrheae and oral commensal Neisseria spp. carry known rRNA methylase genes. Antimicrob Agents Chemother 1999;43:1367 / 72. [10] Kataja J, Huovinen P, Skurnik M, Seppalla H. Erythromycin resistance genes in group A streptococci in Finland. Antimicrob Agents Chemother 1999;43:48 /52. [11] Roberts MC, Sutcliffe J, Courvalin P, Jensen LB, Rood J, Seppalla H. Nomenclature for macrolide and macrolide / lincosamide /streptogramin B resistance determinants. Antimicrob Agents Chemother 1999;43:2823 /30. [12] King A, Bathgate T, Phillips I. Erythromycin susceptibilitiy of viridans streptococci from the normal throat flora of patients treated with azithromycin or clarithromycin. Clin Microbiol Infect 2002;8:85 /92. [13] Perez-Trallero E, Vicente D, Montes M, Marimon JM, Pineiro L. High proportion of pharyngeal carriers of commensal streptococci resistant to erythromycin in Spanish adults. J Antimicrob Chemother 2001;48:225 /9. [14] Rodriguez-Avial I, Rodriguez-Avial C, Culebras E, Benitez A, Picazo JJ. Distribution of mef (A) and erm (B) genes in macrolide resistant blood isolates of viridans group streptococci. J Antimicrob Chemother 2001;47:727 /8. [15] Ono T, Shiota S, Hirota K, Nemoto K, Tsuchiya T, Miyake Y. Susceptibilities of oral and nasal isolates of Streptococcus mitis and Streptococcus oralis to macrolides and PCR detection of resistance genes. Antimicrob Agents Chemother 2000;44:1078 / 80. [16] Jacobs JA, van Baar GJ, London NH, Tjhie JH, Schouls LM, Stobberingh EE. Prevalence of macrolide resistance genes in clinical isolates of the Streptococcus anginosus (‘S. milleri ’) group. Antimicrob Agents Chemother 2001;45:2375 /7. [17] Arpin C, Canron MH, Maugein J, Quentin C. Incidence of mef A and mef E genes in viridans group streptococci. Antimicrob Agents Chemother 1999;43:2335 /6. [18] Syrogiannopoulos GA, Grivea IN, Fitoussi F, Doit C, Katopodis GD, Bingen E, et al. High prevalence of erythromycin resistance of Streptococcus pyogenes in Greek children. Pediatr Infect Dis J 2001;20:863 /8. [19] Tzelepi E, Kouppari G, Mavroidi A, Zaphiropoulou A, Tzouvelekis L. Erythromycin resistance amongst group A beta-haemolytic streptococci isolated in a paediatric hospital in Athens Greece. J Antimicrob Chemother 1999;43:745 /6. [20] Ioannidou S, Tassios PT, Zachariadou L, et al. In vitro activity of telithromycin (HMR 3647) against Greek Streptococcus pyogenes and Streptococcus pneumoniae clinical isolates with different macrolide susceptibility, Clin Microbiol Infect 2003;9:704 /7. [21] Zachariadou L, Papaparaskevas J, Paraskakis I, et al. Predominance of two M-types among erythromycin-resistant group A
S. Ioannidou et al. / International Journal of Antimicrobial Agents 22 (2003) 626 /629 streptococci from Greek children, Clin Microbiol Infect 2003;9:310 /4. [22] Lipsitch M, Samone MH. Antimicrobial use and antimicrobial resistance: a population perspective. Emerg Infect Dis 2002;8:347 /54.
629
[23] Reig M, Galan JC, Baquero F, Perez-Diaz JC. Macrolide resistance in Peptostreptococcus spp. mediated by class A erm : possible source of macrolide /lincosamide /streptogramin B resistance in Streptococcus pyogenes . Antimicrob Agents Chemother 2001;45:630 /2.