Macrolide resistance phenotypes of commensal viridans group streptococci and Gemella spp. and PCR detection of resistance genes

International Journal of Antimicrobial Agents 23 (2004) 582–589

Macrolide resistance phenotypes of commensal viridans group streptococci and Gemella spp. and PCR detection of resistance genes P. Cerdá Zolezzi, M.C. Rubio Calvo, L. Millán, P. Goñi, M. Canales, S. Capilla, E. Durán, R. Gómez-Lus∗ Department of Microbiology, Faculty of Medicine, Clinical University Hospital “Lozano Blesa”, University of Zaragoza, c/Domingo Miral s/n, 50009 Zaragoza, Spain Received 25 June 2003; accepted 29 October 2003

Abstract One hundred and sixty viridans group streptococci (VGS) and 26 Gemella spp. resistant to erythromycin were studied to detect macrolide lincosamide and streptogramin B (MLSB ) phenotypes and to investigate resistance rates to other antibiotics. The M phenotype was most prevalent in both bacterial groups (59.6% in VGS, 69.2% in gemellae) and the iMLSB phenotype was found least often (9.3 and 13.9%, respectively). All isolates with M phenotype had the mef(A/E) gene, being prevalent the mef(E) subclass. cMLSB and iMLSB strains contained the erm(B) gene, alone or in combination with the mef(A/E) gene. Thirteen isolates were intermediately resistant to quinupristin/dalfopristin and 11 strains showed low susceptibility to telithromycin. Linezolid was active against all the isolates tested and tetracycline resistance was the major one in VGS (41.6%) and Gemella spp. (46.2%). © 2004 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. Keywords: Viridans group streptococci; Gemella spp.; Macrolides; Resistance genes; Antibiotic susceptibilities

1. Introduction Viridans group streptococci (VGS) and Gemella spp. are commensal bacteria of the human upper respiratory tract, sharing the habitat with pathogens like Streptococcus pneumoniae and S. pyogenes. Viridans streptococci are also recognized as cause of systemic diseases included bacterial endocarditis, bacteraemia, especially in neutropenic patients, and pneumonia [1,2]. Gemellae, as opportunistic pathogens, are able to cause severe localized and generalized infections [3–5]. Beta-lactam agents are the preferred choice for the treatment and prophylaxis of infection caused by these bacteria, but macrolides and related drugs are recommended for penicillin-resistant viridans streptococci and for alternative treatment in allergic patients [6,7].

∗

Corresponding author. Tel.: +34-976-761692; fax: +34-976-761693. E-mail address: [email protected] (R. G´omez-Lus).

Resistance to macrolides and other related antibiotics has spread among streptococci [8,9]. Two major mechanisms account for resistance to macrolides, lincosamides and streptogramin B antibiotics (MLSB ) in Gram positive bacteria, conferring distinct phenotypes. The first mechanism is mediated by methylation of the ribosomal target of these antibiotics (MLSB resistance) and is encoded by erm genes (erythromycin ribosome methylase) [10]. Expression of MLSB resistance can leads to cross-resistance to macrolides, lincosamides and streptogramins (cMLSB ), or can be inducible (iMLSB ). Inducible expression in streptococci and enterococci gives rise to a variety of phenotypes, including high- or low-level resistance to erythromycin with susceptibility or resistance to clindamycin [11]. The second mechanism is mediated by an active efflux pump, encoded by mef(A/E) gene [12,13]. The mef(A/E) gene causes resistance to 14- and 15-membered macrolides compounds only, and the encoding phenotype is designed M [14]. Two subclasses of the mef(A/E) gene, mef(A) [12], originally found in S. pyogenes, and mef(E) originally found in S. pneumoniae [13], have been described. mef(A) and mef(E) are 90%

0924-8579/$ – see front matter © 2004 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/j.ijantimicag.2003.10.014

P. Cerd´a Zolezzi et al. / International Journal of Antimicrobial Agents 23 (2004) 582–589

identical at the nucleotide level but they are endowed with important genetic differences [15]. During recent years, increased rates of macrolide resistance in clinical isolates of viridans group streptococci have been reported [16–18]. In Spain, some authors found resistance rates to erythromycin of 70.6% in oral viridans group streptococci, being predominantly the M phenotype [19]. As far as we know, there is not information about resistance to erythromycin in Gemella spp. Linkage of multiple antibiotic resistance genes, which implies resistance to various antimicrobials, has been observed. In pneumococci, the most prevalent resistance pattern was resistance to erythromycin, tetracycline and chloramphenicol, although resistance to erythromycin and kanamycin has also been found [20]. The aims of this study were, using strains mainly from normal flora, to determine the macrolide resistance phenotypes and resistance genes in viridans group streptococci as well as in Gemella spp. The resistance rates to other antimicrobial agents such as tetracycline, chloramphenicol, aminoglycosides, quinupristin/dalfopristin and linezolid, and their possible association with erythromycin resistance were also investigated.

2. Materials and methods 2.1. Bacterial strains A total of 160 viridans group streptococci and 26 Gemella spp. resistant to erythromycin and derived from 172 patients were isolated between October 2001 and March 2003 from the Microbiology Service, Lozano Blesa Clinical University Hospital (Zaragoza, Spain). The isolate sources were pharyngeal exudate (85), sputum (51), bronchial aspirate (18), nasal sample (13), oral sample (5) and others (14). Strains were identified on the basis of colony morphology, ␣-haemolysis, optochin susceptibility and Gram stain. Species identification was performed using the API20 Strep System (API System, bioMerieux, Marcy-l’Etoile, France). The isolates were assigned to group/species, according to the criteria of Facklam [21]. 2.2. Detection of MLSB resistance phenotypes Classification of macrolide resistance was based on the method of Seppälä et al. [22]. Briefly, erythromycin (15 ␮g), clindamycin (2 ␮g) (Bio-Rad, La Coquette, France), and miocamycin (Neo-Sensitabs, Taastrup, Denmark) disks were placed 15–20 mm apart on Mueller–Hinton blood agar plate (MHB) (Biomedics, Madrid, Spain) streaked with a 0.5 McFarland standard bacterial suspension. After incubation, a blunting of the clindamycin or miocamycin inhibition zone towards the erythromycin disk was interpreted as iMLSB phenotype. Resistance to clindamycin and miocamycin with no blunting indicated cMLSB phe-

583

notype. M phenotype was characterized by susceptibility to clindamycin and miocamycin with no blunting of the inhibition zone near the erythromycin disk. This phenotype was confirmed by performing the induction test described by Malke [23]. A bacterial suspension was grown in Brain Heart Infusion (Difco, Detroit, MI, USA) for 3 h at 37 ◦ C with 0.05 ␮g/ml of erythromycin, and then, the former disk test was repeated. 2.3. Antimicrobial susceptibility testing Antimicrobial susceptibility testing was performed by a standard agar diffusion test with commercial disks (Bio-Rad, La Coquette, France) and a standard agar dilution method according to the guidelines established by the National Committee for Clinical Laboratory Standards [24,25]. The following antimicrobial agents were used: erythromycin, clindamycin, tetracycline, chloramphenicol, streptomycin, spectinomycin (Sigma, St. Louis, MO, USA), azithromycin (Pfizer, Madrid, Spain), miocamycin (Menarini S.A., Badalona, Spain), telithromycin, quinupristin/dalfopristin (Aventis, Madrid, Spain), kanamycin (Amersham Life Science, China), gentamicin, minocycline and linezolid (Pharmacia & Upjohn Inc., Kalamazoo, MI). S. pneumoniae ATCC 49619 was used as quality control strain. The range of interpretative categories for each antibiotic were those recommended by the NCCLS [26]. The MIC breakpoint for miocamycin resistance was that defined by the Comité de l’Antibiogramme de la Societé Française de Microbiologie [27]. As there are no defined MIC breakpoints for Gemella spp., we used those of VGS, taking account the similarities of these two bacterial genera. 2.4. Detection of erythromycin resistance genes DNA samples of erythromycin-resistant VGS and Gemella spp. were prepared as described by Ausubel et al. [28]. The DNAs of the isolates were amplified with primers specific for the erm(A), erm(B), erm(C), erm(TR), and mef(A/E) genes. PCRs of erm(A), erm(B), erm(C), and mef(A/E) genes were performed with oligonucleotide primers and conditions reported by Sutcliffe et al. [29]. The erm(TR) gene was detected with the primers and conditions described by Seppälä et al. [30]. Amplifications were performed in a Perkin-Elmer Cetus DNA thermal cycler (Perkin-Elmer, Norwalk, CT). PCR products were resolved by electrophoresis on 1.5% agarose gels. In order to discriminate between mef(A) and mef(E), a PCR-restriction fragment length polymorphism analysis was done. The mef(A/E) amplicon was digested with the BamHI restriction enzyme. In the mef(A) amplicon there is one BamHI site, so restriction generates two fragments of 282 and 64 bp, while in the mef(E) amplicon there are no BamHI restriction sites. PCR reagents and BamHI enzyme were purchased from Promega (Promega, Madison, WI, USA).

584

P. Cerd´a Zolezzi et al. / International Journal of Antimicrobial Agents 23 (2004) 582–589

2.5. Tests of statistical significance The χ2 -test was used to attribute statistical significance to differences of resistant rates among the different macrolide-resistance phenotypes and between groups of isolates.

3. Results Of the 160 isolates of VGS examined in this study, 123 were identified as S. mitis, 27 as S. oralis, 7 as S. sanguinis and 3 as S. salivarius. Within Gemella spp., 13 were identified as G. haemolysans and 13 as G. morbillorum (Table 1). 3.1. MLSB resistance phenotypes The isolates were examined for MLSB phenotypes by a modification of the double-disk test. The M phenotype was the most prevalent in VGS streptococci (60.0%) as well as in Gemella spp. (69.2%). Within S. sanguinis only this phenotype was found. All the strains with M phenotype were negative for the induction test. cMLSB was the most common phenotype only for S. oralis (48.2%) and iMLSB occurred least often in all the bacterial species analyzed (Table 1). Eight isolates (five S. mitis, two S. oralis and one S. salivarius) with iMLSB phenotype showed a high level of clindamycin resistance compared with miocamycin. 3.2. Susceptibility to MLSB antibiotics and telithromycin Table 2 shows the mode MIC values and numbers of strains sensitive, intermediate or resistant to macrolides and other antibiotics according to MLSB phenotypes, for erythromycin-resistant VGS and Gemella spp. The MIC range of erythromycin for the tested strains was 0.5 to >128 mg/l, with a MIC50 of 8 mg/l and a MIC90 of >128 mg/l. VGS isolates with M phenotype ranged from 0.5 to 32 mg/l (MIC50 4 mg/l), and Gemella spp. isolates from 0.5 to 8 mg/l (MIC50 4 mg/l). All the strains with this phenotype were susceptible to miocamycin and all but two to

clindamycin (MIC 0.5 mg/l), although they were susceptible by the agar disk test. The MIC50 and MIC90 for miocamycin and clindamycin in cMLSB VGS and Gemella spp. were ≥64 mg/l and all the isolates with this phenotype showed high resistance to erythromycin (MIC 128, >128 mg/l). Heterogeneous susceptibility patterns were observed among isolates of the iMLSB phenotype. In VGS, the MIC range of erythromycin was 2 to >128 mg/l, with a MIC50 of 64 mg/l and a MIC90 of >128 mg/l. Resistance to miocamycin was 13.3% with a MIC50 of 2 mg/l, whereas clindamycin presented a high percentage of resistance (60.0%) with a MIC50 of 4 mg/l, being less active for iMLSB isolates. Only one Gemella sp. (G. morbillorum) presented iMLSB phenotype with a MIC of erythromycin of 4 mg/l and MICs of 2 and 0.5 mg/l for miocamycin and clindamycin, respectively. S. mitis followed a similar pattern to global VGS behaviour for these three antibiotics. S. oralis and S. salivarius with iMLSB phenotype had a MIC50 for erythromycin of 128 mg/ml, with high MICs values and high resistance rates to miocamycin and clindamycin. There were no important differences between G. haemolysans and G. morbillorum. Azithromycin MIC values were similar to those of erythromycin. However, in VGS, the MIC50 of azithromycin, except in cMLSB phenotype, was two-fold greater. The MIC range of telithromycin for VGS was ≤0.03 to 2 mg/l (MIC50 , 0.12 mg/l; MIC90 , 0.5 mg/l). Only one strain with the cMLSB phenotype was resistant to this antibiotic (MIC, 2 mg/l). Nine strains were intermediately resistant to telithromycin and five of them displayed the former phenotype (P < 0.05). All but one Gemella sp. were susceptible. Telithromycin intermediately resistant G. morbillorum also showed cMLSB phenotype. The MIC50 and MIC90 of quinupristin/dalfopristin for VGS were 1 and 2 mg/l respectively, whereas these MIC values in Gemella spp. were two-fold lower, with higher values for cMLSB and iMLSB phenotypes. Intermediate resistance rates to quinupristin/dalfopristin of 7.5% in VGS and 7.7% in Gemella spp. were observed. These strains showed mainly cMLSB phenotype, with significant differences (P < 0.01) between cMLSB and M phenotype for VGS. 3.3. Susceptibility to other antibiotics

Table 1 Rates of MLSB resistance phenotypes among viridans group streptococci and Gemella spp. Bacterial species (no strains)

Viridans group S. mitis (123) S. oralis (27) S. sanguinis (7) S. salivarius (3) Gemella spp. G. haemolysans (13) G. morbilorum (13)

Macrolides phenotypes (%) M

cMLSB

iMLSB

60.0 61.8 44.4 100 33.3

30.6 28.5 48.2 0 33.3

9.4 9.8 7.4 0 33.3

69.2 76.9 61.5

26.9 23.1 30.8

13.9 0 7.7

As shown in Table 2, resistance to other antimicrobial agents was the major pattern for cMLSB and iMLSB phenotypes, whereas the greatest percentage of M strains (except for S. oralis), only showed resistance to 14- and 15-membered macrolides (52.4% in VGS and 66.6% in Gemella spp.). Resistance (including intermediately resistant strains) to tetracycline (and minocycline) was the most important one in VGS (overall 41.6%) as well as in Gemella spp. (overall 46.2%) and was found mainly associated with cMLSB and iMLSB phenotypes (P < 0.005). Tetracycline MIC50 modes in VGS and in Gemella spp. were 0.25 mg/l for the M phenotype and 16 mg/l for cMLSB and iMLSB phenotypes.

P. Cerd´a Zolezzi et al. / International Journal of Antimicrobial Agents 23 (2004) 582–589

585

Table 2 Mode MICs (mg/l) and number of strains sensitive/intermediate or resistant to antibiotics listed M (96) Mode

Constitutive MLS (49) n

Mode

S

I

R

Inducible MLS (15)

n

Mode

S

I

R

n S

I

R

Viridans group streptococci (n = 160) Ery Azi Mio Cli Q/D Tel Tet Mino Clm Kan Lin

8 16 0.12 ≤0.06 1 0.12 0.25 ≤0.12 2 64 1

96 0 2 96 95 93 93 82 83 93 95 96

2 2 0 1 3 3 3 2 2 0 0

94 92 0 0 0 0 11 11 1 1 0

>128 >128 >64 >64 1 0.06 16 16 4 64 1

49 0 0 0 0 41 42 10 10 46 47 49

0 0 0 0 8 6 1 3 2 0 0

49 49 49 49 0 1 38 36 1 2 0

>128 >128 2 ≤0.06 1 0.06 16 16 2 64 1

15 0 0 13 4 14 15 2 2 14 14 15

0 0 0 2 1 0 1 1 0 0 0

15 15 2 9 0 0 12 12 1 1 0

Gemella spp. (n = 26) Ery Azi Mio Cli Q/D Tel Tet Mino Clm Kan Lin

8 8 0.12 ≤0.06 1 0.12 0.25 0.12 2 64 1

18 0 1 18 18 17 18 13 13 18 18 18

2 1 0 0 1 0 0 1 0 0 0

16 16 0 0 0 0 5 4 0 0 0

>128 >128 >64 >64 0.5 ≤0.03 16 16 2 64 1

7 0 0 0 0 6 6 1 1 7 7 7

0 0 0 0 1 1 0 0 0 0 0

7 7 7 7 0 0 6 6 0 0 0

4 4 2 0.5 1 0.06 16 16 2 64 0.25

1 0 0 1 0 1 1 0 0 1 1 1

0 0 0 1 0 0 0 0 0 0 0

1 1 0 0 0 0 1 1 0 0 0

Ery: erythromycin; Azi: azithromycin; Mio: miocamycin; Cli: clindamycin; Q/D: quinupristin/dalfopristin; Tel: telithromycin; Tet: tetracycline; Mino: minocycline; Clm: chloramphenicol; Kan: kanamycin; Lin: linezolid.

Of the species tested S. salivarius showed the highest resistance rate (66.7%; MIC50 , 16 mg/l) and S. sanguinis, that only displayed the M phenotype, the lowest (28.6%; MIC50 , 0.5 mg/l). G. morbillorum with the cMLSB phenotype were completely susceptible to tetracycline. Similar values were found for minocycline. The range of MIC for chloramphenicol in VGS was 1–16 mg/l (MIC50 , 2 mg/l and MIC90 , 4 mg/l). The resistance (including intermediately resistant strains) rate was 4.4% and all strains with this characteristic were S. mitis. There were no significant differences in resistance rates to chloramphenicol between macrolide resistance phenotypes (P > 0.5). All Gemella spp. isolates were susceptible to chloramphenicol. We also analyzed the presence of isolates with high-level aminoglycoside resistance (MIC, ≥512 mg/l). Two S. mitis (M and iMLSB phenotypes) and two S. oralis (cMLSB phenotype) showed high-level kanamycin resistance. S. oralis displayed the highest MIC values and a mode of 64 mg/l. Four S. mitis (two with the cMLSB phenotype and two with the iMLSB phenotype) and two S. oralis (cMLSB phenotype) exhibited high-level resistance to streptomycin, with a MIC of >1024 mg/l, but were susceptible to spectinomycin. No isolates were highly resistant to gentamicin. All erythromycin-resistant and intermediately resistant strains were susceptible to linezolid and were inhibited at

2 mg/l or less of the new oxazolidinone (MIC50 , 1 mg/l and MIC90 , 2 mg/l). 3.4. Erythromycin resistance determinants The presence of the erythromycin resistance genes erm(A), erm(B), erm(C), erm(TR), and mef(A/E) was investigated by PCR. All the isolates with M phenotype gave the expected PCR product (346 bp) of the mef(A/E) gene. The erm(B) gene was detected in all the strains with cMLSB or iMLSB phenotype, either alone or in combination with the mef(A/E) gene in 48.2% of the isolates with the former phenotype and in 56.2% with the latter one. None of the isolates contained the erm(A), erm(C) or erm(TR) genes. The mef(A/E) amplicons were analyzed by restriction to differentiate mef(A) and mef(E). mef(E) was predominant in VGS (95.8%) as well as in gemellae (95.2%). Only five VGS (two S. mitis, two S. oralis, and one S. sanguinis) and one G. haemolysans, carried the mef(A) subclass.

4. Discussion We studied the distribution of MLSB resistance phenotypes among VGS and Gemella spp., mainly isolated from the upper respiratory tract, and its relationship with resistance to other classes of antimicrobial agents.

586

P. Cerd´a Zolezzi et al. / International Journal of Antimicrobial Agents 23 (2004) 582–589

As previously reported, S. mitis was the most prevalent specie isolated [18,31,32]. Determination of MLSB phenotypes showed that the M phenotype occurred most frequently in VGS (59.6%) as well as in Gemella spp. (69.2%), with inducible resistance being found in only 9.3 and 13.9% respectively. These results agree with those found in VGS isolated from the oropharynx in Spain by Aracil et al. [19] and in Greece by Ioannidou et al. [18]. Notably, in blood VGS from Spain, cMLSB phenotype was documented to most important (56%), and no strains displayed the iMLSB phenotype [32,33]. We only looked at two VGS from blood (data not shown), one S. sanguinis with the M phenotype and one S. mitis with the iMLSB phenotype. In France, Arpin et al. [34] also found a high rate of VGS with the cMLSB phenotype (77.1%) compared with the M phenotype (22.9%), and they attributed this to the lower incidence of ␤-haemolytic streptococci and pneumococci with mef(A/E) genes in France. In our environment, although the cMLSB phenotype is predominant in pneumococci [20], we did not find the same in the VGS and gemellae analyzed. A study of the SENTRY Antimicrobial Surveillance Program [35] which investigated antimicrobial susceptibility patterns of clinical isolates of VGS from various regions in the world, showed that the M phenotype was significantly more prevalent in the Americas compared to the Asia-Pacific or Europe, where cross resistance to MLS antibiotics was higher. The MIC values of macrolides and clindamycin in VGS agree with those found by some authors [33]. Lower MIC values of erythromycin, ranged from 0.5 to 32 mg/l, with susceptibility to clindamycin and miocamycin, were found in M phenotype strains. cMLSB phenotype isolates showed a high level of erythromycin resistance with cross-resistance to clindamycin and miocamicyn, exhibiting MIC values of >64 mg/l. The high heterogeneity found in iMLSB phenotype can be explained in part by the fact that the patterns of macrolide inducers depend on the structure of the attenuator controlling the gene expression, and on the genetic background or bacterial host, which leads in the case of streptococci, to a large variety of phenotypes, including high or low-level resistance to erythromycin with susceptibility or resistance to clindamycin. Most members of the MLSB group, including clindamycin and 16-membered-macrolides, are inducers at various degrees of ErmB methylase production [11]. With respect to differences between species/groups, S. oralis was the only one that showed prevalence of cMLSB phenotype and exhibited the highest resistance rates to MLSB antibiotics. In contrast, S. sanguinis only displayed the M phenotype with the lowest resistance to these antibiotics. Rodriguez-Avial et al. [33] also found that the prevalence of M phenotype among S. sanguinis isolated from blood was statistically significant (P < 0.01). MIC values of MLSB antibiotics in Gemella spp. were similar to those found in VGS, with a narrow range of erythromycin values in M phenotype strains.

To clarify the mechanisms of resistance, PCR amplification of macrolide-resistance genes was performed. Among isolates with cMLSB or iMLSB phenotype, the erm(B) gene was always detected, which agrees with previous studies [32,36]. Jacobs et al. [37] found the erm(TR) gene in two clinical isolates of S. anginosus. None of the strains analyzed in this study possessed the erm(A), erm(C) or erm(TR) gene. All M phenotype isolates harboured the mef(A/E) gene, as reported by others authors [9,34,36–38]. The mef(E) was the predominant subclass, as previously observed for S. pneumoniae [39]. Most of the authors did not differentiate the two mef(A/E) subclasses. In VGS we only found one report of Arpin et al. [34] which documented the prevalence of mef(E) in clinical strains. We considered that the differentiation of the two subclasses of mef(A/E) gene is of concern because they are endowed with important genetic differences and give information about the possible origin of these genes and about the gene transfer with others bacteria. The presence of the erm(B) gene and the prevalence of mef(E) subclass with the absence of erm(A), erm(C) and erm(TR) genes, suggested that VGS and gemellae could exchange genetic material especially with S. pneumoniae. To our knowledge, this is the first description of the erm(B), mef(E) and mef(A) genes in the genus Gemella. In previous reports [18,35,40], quinupristin/dalfopristin demonstrated good in vitro activities against streptococci, although different breakpoints values were used. Alcaide et al. [6] found that MICs values for highly erythromycin-resistant VGS were higher (MIC90 , 8 mg/l) than those found in erythromycin-intermediate resistant and susceptible strains (MIC90 , 1 mg/l). In our study, quinupristin/dalfopristin showed better in vitro activity against strains with M phenotype than against isolates with cMLSB and iMLSB phenotypes, with significant statistical differences in resistance rates among macrolides phenotypes for VGS (P < 0.01). The same is true for the ketolide telithromycin; most of the intermediate strains and the single resistant strain belonged to the cMLSB and iMLSB phenotypes. The resistance rate to telithromycin in cMLSB phenotype was significantly different (P < 0.05) from that found in the M phenotype VGS strains. However, telithromycin was one of the most active antimicrobial agents tested, with the lowest MIC50 and MIC90 values in VGS and in Gemella spp. This result agrees with those documented by Alcaide et al. [41] and Casellas et al. [42], although they did not find differences in resistance rates between macrolide phenotypes. The resistance rate of tetracycline in VGS was higher (41.6%) than those reported by Ioannidou et al. [18] (17%), and was mainly associated with cMLSB and iMLSB phenotypes (P < 0.005); probably due to the co-localization of erythromycin (erm(B)) and tetracycline (tet(M)) resistance genes on streptococcal transposons [43]. In contrast, Ioannidou et al. [18] found that tetracycline resistance was equally distributed between M and cMLSB resistance phenotypes. In pneumococci, Seral et al. [20] reported a tetracycline resistance rate of 82% among erythromycin-resistant

P. Cerd´a Zolezzi et al. / International Journal of Antimicrobial Agents 23 (2004) 582–589

isolates, and found a strong association of erm(B) and tet(M) genes with Tn1545-related elements. They also found association percentages of 50.6 and 11.8% of erm(B) and mef(A) genes with chloramphenicol resistance (catpC194 determinant) respectively. We did not find significant differences in resistance rates of chloramphenicol among macrolide resistance phenotypes (P > 0.5). All chloramphenicol-resistant strains (4.4%) were S. mitis. The MIC90 was 4 mg/l, two-fold lower than that found by Alcaide et al. [41] in VGS isolated from blood of neutropenic patients with cancer. High-level kanamycin (2.4%) and streptomycin (3.6%) resistance was detected in our study but only documented in S. mitis and S. oralis. All but one displayed cMLSB or iMLSB phenotype although there were not significant differences among phenotypes. Streptomycin-resistant strains were susceptible to spectinomycin, suggesting that this resistance could be due to the presence of the aminoglycoside modifying enzyme ANT(6) or to chromosomal mutations [44,45]. In a study performed in 1982 [46] with invasive VGS, high-level kanamycin resistance in four S. mitis and one S. sanguinis was associated to MLSB and tetracycline resistance, and in three S. mitis and in one S. sanguinis to streptomycin resistance. We found combined streptomycin and kanamycin resistance in two strains due to the presence of ant(6) and aph(3 )-III genetic determinants (data not shown). Farber et al. [47] also found two S. mitis with both determinants. High-level gentamicin resistance was not documented neither in our study, or in that of Iaonnadou et al. [18], although a study from South Africa [48], revealed the presence of two S. mitis with chromosomally mediated high-level gentamicin resistance. Linezolid was reported to have excellent and nearly complete in vitro activity against Gram-positive cocci [35,49]. In the current report, linezolid demonstrated the best activity of all the antimicrobial agents analyzed, being susceptible 100% of the isolates, with no differences in MIC values among MLSB phenotypes. As far as we know, there is little information about antibiotic resistance in Gemella spp. Most of the infections caused by this bacterial genus are successfully treated with benzylpenicillin or amoxicillin associated with gentamicin [5,7]. The only report of antimicrobial susceptibility found [50] analyzed five strains of G. haemolysans, which was susceptible to penicillin G, clindamycin and chloramphenicol. One of them showed resistance to tetracycline and erythromycin and all of them displayed low level resistance to aminoglycosides. The pattern found by these authors, like ours, was comparable with that of streptococci. VGS and Gemella spp. are normal commensals of the upper respiratory tract. The erythromycin resistance rate in ␣-haemolytic bacteria was 78.9% (data not shown), which is one of the highest values found in the literature [6,16,18,19,35]. The surveillance of resistance to an-

587

timicrobials in VGS and gemellae is of concern because these bacterial groups offer a pool of genetic material which may be exchanged with other bacteria, including pathogenic species that share their habitats and because of the serious invasive diseases they can cause. Besides, the selective pressure exerted by the antibiotic use could select strains with resistance to more than one antimicrobial agent.

Acknowledgements This study was supported in part by Ministerio de Sanidad y Consumo grant (Project FIS 01/0210), and Departamento de Educación y Cultura del Gobierno Autónomo de Aragón grant (Project DGA P056/2001). P. Cerdá was the recipient of fellowship B102/2003 from Diputación General de Aragón, Departamento de Educación y Ciencia.

References [1] Kennedy HF, Gemmell CG, Bagg J, Gibson BES, Michie JR. Antimicrobial susceptibility of blood cultures isolates of viridans group streptococci: relationship to a change in empirical antibiotic therapy in febrile neutropenia. J Antimicrob Chemother 2001;47:693–6. [2] Douglas CWI, Heath J, Hampton K, Preston FE. Identity of viridans streptococci isolated from cases of infective endocarditis. J Med Microbiol 1993;39:179–82. [3] Fresard A, Michel VP, Rueda X, Aubert G, Dorche G, Lucht F. Gemella haemolysans endocarditis. Clin Infect Dis 1993;16:586–7. [4] Omran Y, Wood CA. Endovascular infection and septic arthritis caused by Gemella morbillorum. Diagn Microbiol Infect Dis 1993;16:131–4. [5] La Scola B, Didier R. Molecular identification of Gemella species from three patients with endocarditis. J Clin Microbiol 1998;36:866– 71. [6] Alcaide F, Catarrala J, Linares J, Gudiol F, Martin R. In vitro activities of eight macrolide antibiotics and RP-59500 (quinupristin/dalfopristin) against viridans group streptococci isolated from blood of neutropenic cancer patients. Antimicrob Agents Chemother 1996;40:2117–20. [7] Akiyama K, Taniyasu N, Hirota J, Iba Y, Maisawa K. Recurrent aortic valve endocarditis caused by Gemella morbillorum. Report of a case and review of the literature. Jpn Circ J 2001;65:997–1000. [8] Goldstein FW, Acar JF. the Alexander Project Collaborative Group. Antimicrobial resistance among lower respiratory tract isolates of Streptococcus pneumoniae: results of a 1992–93 western Europe and USA collaborative surveillance study. The Alexander Project Collaborative Group. J Antimicrob Chemother 1996;38:71–84. [9] 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. [10] Leclerq R, Courvalin P. Bacterial resistance to macrolides, lincosamide and streptogramin antibiotics by target modification. Antimicrob Agents Chemother 1991;35:1267–72. [11] Leclercq R. Mechanisms of resistance to macrolides and lincosamides: nature of the resistance elements and their clinical implications. Clin Infect Dis 2002;34:482–92. [12] Clancy J, Petitpas J, Did-Hajj F, et al. Molecular cloning and functional analysis of a novel macrolide-resistance determinant, mefA, from Streptococcus pyogenes. Mol Microbiol 1996;22:867–79.

588

P. Cerd´a Zolezzi et al. / International Journal of Antimicrobial Agents 23 (2004) 582–589

[13] Tait-Kamradt A, Clancy J, Cronan M, et al. mefE is necessary for the erythromycin-resistance M phenotype in Streptococcus pneumoniae. Antimicrob Agents Chemother 1997;41:2251–5. [14] Sutcliffe J, Tait-Kamradt A, Wondrack L. Streptococcus pneumoniae and Streptococcus pyogenes resistant to macrolides but sensitive to clindamycin: a common resistance pattern mediated by an efflux system. Antimicrob Agents Chemother 1996;40:1817–24. [15] Del Grosso M, Iannelli F, Messina C, et al. Macrolide efflux genes mef(A) and mef(E) are carried by different genetic elements in Streptococcus pneumoniae. J Clin Microbiol 2002;40:774–8. [16] Doern GV, Ferraro MJ, Brueggemann AB, Ruolff KL. Emergence of high rates of antimicrobial resistance among viridans group streptococci in US. Antimicrob Agents Chemother 1996;40:891–4. [17] Teng LJ, Hsueh PR, Chen YC, Ho SW, Luh KT. Antimicrobial susceptibility of viridans group streptococci in Taiwan with an emphasis on the high rates of resistance to penicillin and macrolides in Streptococcus oralis. J Antimicrob Chemother 1998;41: 621–7. [18] Ioannidou S, Tassios PT, Kotsovili-Tseleni A, Foustoukou M, Legakis NJ, Vatopoulus A. 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. [19] Aracil B, Miñambres M, Oteo J, Torres C, Gómez-Garcés JL, Alós 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. [20] Seral C, Castillo FJ, Rubio-Calvo MC, Fenoll A, Garc´ıa C, Gómez-Lus R. Distribution of resistance genes tet(M), aph3 -III, catpc194 and the integrase gene of Tn1545 in clinical Streptococcus pneumoniae harbouring erm(B) and mef(A) genes in Spain. J Antimicrob Chemother 2001;47:863–6. [21] Facklam R. What happened to the streptococci: overview of taxonomic and nomenclature changes. Clin Microbiol Rev 2002;15:613– 30. [22] Seppälä H, Nissinen A, Yu Q, Huovinen P. Three different phenotypes of erithromycin-resistant Streptococcus pyogenes in Finland. J Antimicrob Chemother 1993;32:885–91. [23] Malke H. Zonal-pattern resistance to lincomycin in Streptococcus pyogenes: genetic and physical studies. In: Schlessinger D, editor. Microbiology. Washington, DC: American Society for Microbiology; 1982. p. 142–5. [24] National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial disk susceptibility test. 6th ed. Approved standard M2-A6. Wayne, PA: National Committee for Clinical Laboratory Standards; 1997. [25] National Committee for Clinical Laboratory Standards. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. 5th ed. Approved standard M7-A5. Wayne, PA: National Committee for Clinical Laboratory Standards; 2000. [26] National Committee for Clinical Laboratory Standards. Performance standards for susceptibility testing (1982): tenth informational supplement M100-S10 (M7). Wayne, PA: National Committee for Clinical Laboratory Standards; 2000. [27] Comité de l’Antibiogramme de la Societé Française de Microbiologie. Communiqué 1999. Par´ıs: Societé Francaise de Microbiologie; 1999. [28] Ausubel FM, Brent R, Kingstron E, Moore DD, Seidman JG, Smith JA. In: Struhl K, editor. Currents protocols in molecular biology. New York: John Wiley & Sons Inc.; 1987. [29] Sutcliffe J, Grebe T, Tait-Kamradt A, Wondrack L. Detection of erythromycin-resistant determinants by PCR. Antimicrob Agents Chemother 1996;40:2562–6. [30] Seppälä H, Skurnik M, Soini H, Roberts MC, Huovinen P. A novel erythromycin resistance methylase gene (ermTR) in Streptococcus pyogenes. Antimicrob Agents Chemother 1998;42:257–62.

[31] De Azavedo JCS, Trpeski L, Pong-Porter S, Matsumura S. The Canadian Bacterial Surveillance Network, Low DE. In vitro activities of fluoroquinolones against antibiotic-resistant blood culture isolates of viridans group streptococci from across Canada. Antimicrob Agents Chemother 1999;43:2299–301. [32] Rodriguez-Avial I, Rodriguez-Avial C, Culebras E, Ben´ıtez 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. [33] Rodriguez-Avial C, Garc´ıa MM, Rodriguez-Avial I, Picazo JJ. Fenotipos de resistencia a macrólidos, lincosamidas y estreptograminas en Streptococcus del grupo viridans aislados de hemocultivos. Rev Esp Quimioter 1999;12:346–51. [34] Arpin C, Canron MH, Maugein J, Quentin C. Incidence of mefA and mefE genes in viridans group streptococci. Antimicrob Agents Chemother 1999;43:2335–6. [35] Gordon KA, Beach ML, Biedenbach DJ, Jones RN, Rhomberg PR, Mutnick AH. Antimicrobial susceptibility patterns of ␤-hemolytic and viridans group streptococci: report from the SENTRY Antimicrobial Surveillance Program (1997–2000). Diagn Microbiol Infect Dis 2002;43:157–62. [36] Poutanen SM, De Azavedo J, Willey B, Low DE, MacDonald KS. Molecular characterization of multidrug resistance in Streptococcus mitis. Antimicrob Agents Chemother; 1999. p. 1505–7. [37] Jacobs JA, van Barr GJ, London NHHJ, et al. Prevalence of macrolide resistance genes in clinical isolates of the Streptococcus anginosus (‘S. milleri’) group. Antimicrob Agents Chemother 2001;45: 2375–7. [38] Luna VA, Coates P, Eady EA, Cove JH, Nguyen TT, Roberts MC. A variety of Gram-positive bacteria carry mobile mef genes. J Antimicrob Chemother 1999;44:19–25. [39] Johnston NJ, De Azavedo C, Kellner JD, Low DE. Prevalence and characterization of the mechanisms of macrolide, lincosamide, and streptogramin in resistance isolates of Streptococcus pneumoniae. Antimicrob Agents Chemother 1998;42:2425–6. [40] Betriu C, Redondo M, Palau ML, et al. Comparative in vitro activities of linezolid, quinupristin-dalfopristin, moxifloxacin, and trovafloxacin against erythromycin-susceptible and -resistant streptococci. Antimicrob Agents Chemother 2000;44:1838–41. [41] Alcaide F, Ben´ıtez MA, Catarrala J, Gudiol F, Liñares J, Mart´ın R. In vitro activities of the new ketolide HMR 3647 (telithromycin) in comparison with those of eight other antibiotics against viridans group streptococci isolated from blood of neutropenic patients with cancer. Antimicrob Agents Chemother 2001;45: 624–6. [42] Casellas JM, Visser M, Mac Dougall N. Estudio multicéntrico en el cono sur americano de la actividad comparativa in vitro de telitromicina en cepas con fenotipos definidos de resistencia asociadas a infecciones respiratorias adquiridas en la comunidad. Rev Esp Quimioter 2001;14:269–74. [43] Rice L. Tn916 family conjugative transposons and dissemination of antimicrobial resistance determinants. Antimicrob Agents Chemother 1998;42:1871–7. [44] Clermont D, Horaud T. Identification of chromosomal antibiotic resistance genes in Streptococcus anginosus (‘S. milleri’). Antimicrob Agents Chemother 1990;34:1685–90. [45] Ounissi H, Derlot E, Carlier C, Courvalin P. Gene homogeneity for aminoglycoside-modifying enzymes in Gram-positive cocci. Antimicrob Agents Chemother 1990;34:2164–8. [46] Horodniceanu T, Buu-Ho¨ı A, Delbos F, Bieth G. High-level aminoglycoside resistance in group A, B, G, D (Streptococcus bovis), and viridans group streptococci. Antimicrob Agents Chemother 1982;21:176–9. [47] Farber BF, Yee Y. High-level aminoglycoside resistance mediated by aminoglycoside-modifying enzymes among viridans streptococci. J Infect Dis 1987;155:948–53.

P. Cerd´a Zolezzi et al. / International Journal of Antimicrobial Agents 23 (2004) 582–589 [48] Kaufhold A, Potgieter E. Chromosomally mediated high-level gentamicin resistance in Streptococcus mitis. Antimicrob Agents Chemother 1993;37:2740–2. [49] Mutnick AH, Biedenbach DJ, Turnidge JD, Jones RN. Spectrum and potency evaluation of a new oxazolidinone, linezolid: report

589

from the SENTRY Antimicrobial Surveillance Program, 1998–2000. Diagn Microbiol Infect Dis 2002;43:65–73. [50] Buu-Ho¨ı A, Sapoetra A, Branger C, Acar JF. Antimicrobial susceptibility of Gemella haemolysans isolated from patients with subacute endocarditis. Eur J Clin Microbiol 1982;1:102–6.