Conjugative transfer frequencies of mef(A)-containing Tn1207.3 to macrolide-susceptible Streptococcus pyogenes belonging to different emm types.

UNLABELLED: The aim of this study was to examine the gene transfer potential of mef(A)-containing Tn120.3 to macrolide-susceptible Streptococcus pyogenes belonging to different emm types. Using the filter mating technique, Tn1207.3 was transferred by conjugation to 23 macrolide-susceptible recipients representing 11 emm types. PCR analysis confirmed the presence of the mef(A) gene and the comEC junction regions of the Tn1207.3 insertion in resultant transconjugants. Significant variation was found in the transfer frequency of Tn1207.3 to different Strep. pyogenes strains, and this phenomenon may contribute to the differences in mef(A) frequency observed among clinical isolates. SIGNIFICANCE AND IMPACT OF THE STUDY: The spread of antimicrobial resistance among pathogenic bacteria is an important problem, but the mechanisms of horizontal transfer between strains and species are often poorly understood. For instance, little is known on how macrolide resistance spreads between strains of the human pathogen Strep. pyogenes and why certain strains more commonly display resistance than others. Here, we show that Strep. pyogenes strains vary greatly in their ability to acquire a transposon encoding macrolide resistance by horizontal gene transfer in vitro. These data provide a novel insight into the transfer of antibiotic resistance between bacterial strains and offer an explanation for the differences in the frequency of resistance determinates and resistance seen among clinical isolates.

Introduction

Streptococcus pyogenes (group A streptococci/GAS) remains an important human pathogen responsible for a wide variety of invasive and noninvasive infections including pharyngotonsillitis, skin and soft tissue infections and bacteraemia. Macrolides are widely used to treat GAS infections in patients allergic to β-lactam antibiotics. The development and spread of macrolide resistance among GAS throughout the world is therefore a major concern.

The drug efflux pump, encoded by the mef gene (currently represented by four subclasses A, E, I and O), is one of the most common mechanisms responsible for resistance to 14- and 15-membered macrolides in GAS and other species (Sutcliffe et al. 1996a; Varaldo et al. 2009). At least three mobile elements carrying the mef(A) gene have been identified in GAS: the Tn1207.3 transposon (Santagati et al. 2003), the Φ10394.4 phage chimera (Banks et al. 2003) and the Φm46.1 element (Brenciani et al. 2010). Studies indicate a significant prevalence of the mef(A) gene among GAS isolates in Europe and elsewhere. For instance, nationwide surveillance in Germany noted a frequency of 31% (Bley et al. 2011), while an Italian study found the prevalence of mef(A) among GAS to be 16% (Creti et al. 2005). Interestingly, there appears to be a skewed distribution of macrolide resistance genes among different GAS strains belonging to different emm types. The mef(A) gene is most frequently harboured by certain emm types such as emm1, emm2, emm3, emm4, emm9, emm12 and emm75, whereas emm types such as emm22, emm77, emm87 and emm89 have not yet been documented to carry any mef determinant (Creti et al. 2005; Grivea et al. 2006; Silva-Costa et al. 2008; Wajima et al. 2013). These observations suggest differences in the capacity of GAS emm types to acquire mef genes, and while the horizontal gene transfer of mef(A)-bearing genetic elements has been demonstrated in vitro by conjugation (Giovanetti et al. 2003; Santagati et al. 2003) and by transduction (Di Luca et al. 2010), no comparison has yet been made of mef gene transfer frequencies among different GAS strains.

The aim of this study therefore was to examine the transfer frequency of mef(A)-containing Tn1207.3 to diverse macrolide-susceptible recipients representing a variety of clinically important emm types.

Results and discussion

Erythromycin-resistant transconjugants were detected in 21/23 of the tested recipient strains, with mean frequencies of transconjugants ranging from 7·20 × 10−8 to 1·13 × 10−6 conjugant/recipient (Table1). Statistical analysis showed that there was a significant difference in the conjugation frequency between strains (anova, P < 0·001). PCR confirmed that all 85 randomly selected transconjugants were positive for the mef(A) gene as well as the 5′ and 3′ junctions of Tn1207.3-comEC. All amplicons were of the expected size. Thus, growth on erythromycin-selective plates can be taken as being strongly indicative of mef(A) acquisition by the previously susceptible, mefA-negative recipient strains and the integration of Tn1207.3 into the comEC locus.

Table 1

Strain characteristics and gene transfer frequencies of the mef(A)-bearing Tn1207.3 element (SMH036) to macrolide-susceptible Strep. pyogenes

emm type	Recipient	Multilocus sequence type	Site of isolation	Mean frequencies of three transfers per recipient ± SEM
emm1	R366	ST28	Blood	9·83 × 10−6 ± 1·87 × 10−6
	R203	ST28	Blood	1·13 × 10−6 ± 6·46 × 10−7
	R057	ST28	Eye	3·86 × 10−6 ± 1·91 × 10−6
	R502	Not available	Blood	4·23 × 10−6 ± 1·85 × 10−6
emm2	R837	ST55	Blood	3·17 × 10−6 ± 3·52 × 10−7
	R007	ST55	Blood	2·37 × 10−6 ± 5·30 × 10−7
emm3	R032	ST315	Throat	1·20 × 10−7 ± 1·52 × 10−8
	R079	ST315	Throat	9·80 × 10−7 ± 1·72 × 10−7
emm4	R317	ST39	Blood	4·50 × 10−7 ± 4·93 × 10−8
	R211	ST39	Blood	6·47 × 10−7 ± 1·05 × 10−7
	R101	ST39	Ear	6·17 × 10−6 ± 3·75 × 10−7
emm9	R480	ST75	Blood	9·20 × 10−7 ± 1·89 × 10−7
	R097	ST75	Pus	1·60 × 10−7 ± 1·53 × 10−8
emm11	R098	ST22	Blood	8·13 × 10−7 ± 1·67 × 10−7
	R693	ST22	Blood	3·67 × 10−7 ± 1·86 × 10−8
emm12	R214	ST36	Blood	1·19 × 10−7 ± 5·01 × 10−8
	R113	ST36	Throat	3·97 × 10−7 ± 4·97 × 10−8
	R044	ST36	Throat	6·00 × 10−7 ± 8·88 × 10−8
emm59	R205	Not available	Blood	4·36 × 10−8 ± 8·60 × 10−9
emm77	R054	ST77	Skin	<3·77 × 10−8 (none detected)
	R115	ST77	Vagina	<1·08 × 10−8 (none detected)
emm81	R202	ST117	Blood	3·97 × 10−7 ± 1·48 × 10−7
emm87	R208	ST62	Blood	7·20 × 10−8 ± 8·99 × 10−9

The association of emm type with conjugative frequency is difficult to assess given the number of isolates of each emm type was low. However, the data show a trend for higher frequencies of conjugation among emm1 and emm4, seen in the region of 10−6 to 10−7 with lower rates seen particularly with the single isolates of emm59 and emm87, in the region of 10−9, Table1. These patterns correlate well with the observed frequency of mef(A) among clinical isolates belonging to these emm types, mef(A) being highly prevalent among emm1 and emm4 isolates but having not been reported among emm59 and emm87 isolates (Creti et al. 2005; Grivea et al. 2006; Silva-Costa et al. 2008; Wajima et al. 2013). These data also highlight the potential for the emergence of mef(A)-mediated resistance in previously susceptible emm types. Two isolates, both belonging to emm77, did not produce transconjugants above the lower limit of detection, Table1, suggesting that a low capacity for acquisition exists among emm77. This again correlates well with epidemiological data, and emm77 has not previously been associated with mef(A). Interestingly, previous work using transduction also failed to transfer mefA into an emm 77 isolate (Di Luca et al. 2010). All seven emm types included here and reported previously to harbour the mef(A) gene were able to acquire it by conjugation in vitro. Taking only the emm types for which at least three isolates were included, emm1, emm4 and emm12, the higher mean frequencies of transconjugants seen with emm1 and emm4 isolates were statistically significant compared to those seen with emm12 isolates. This suggests that an association exists between emm type and the transfer frequency of Tn1207.3 but the testing of a larger collection of isolates is needed to explore this further.

Together the results demonstrate that significant variability exists in mef(A) gene conjugative transfer rates between Strep. pyogenes strains and that this phenomenon may contribute, at least in part, to the different frequencies of mef(A) observed among clinical GAS strains. The mechanisms responsible for these differing rates of mefA transfer are not known, but M proteins have previously been proposed to act as barriers for horizontal gene exchange and thus may play a role (Schmitz et al. 2003). Furthermore, we demonstrate potential for the development of mef(A)-mediated erythromycin resistance in currently susceptible lineages.

Materials and methods

Bacterial isolates

Group A streptococci isolates were collected, as part of a separate study, from two London Hospitals, St. Mary’s and Paddington and Hammersmith, between 1993 and 2005 (McGregor and Spratt 2005). mef(A) recipient strains belonged to the following eleven emm types: emm1, emm2, emm3, emm4, emm9, emm11, emm12, emm59, emm77, emm81 and emm89. Strains were chosen based on their availability for study and because they include some of the most common lineages among clinical disease isolates in Europe (Steer et al. 2009) (Table1). They also include lineages with observed differences in their frequency of mef(A) prevalence (Creti et al. 2005; Grivea et al. 2006; Silva-Costa et al. 2008; Wajima et al. 2013). Isolate SMH036 (emm75) was used as the Tn1207.3 donor and carries the Tn1207.3 element inserted into the comEC locus. Discrimination of mef gene subclasses A, E, O and I by RFLP analysis confirmed the presence of only mef(A) in SMH036. SMH036 was confirmed to be resistant to erythromycin with a MIC of 8 mg l−1 and grew readily on 2 mg l−1 erythromycin-selective blood plates. It was, however, susceptible to fusidic acid and rifampicin. All recipient strains were confirmed to be susceptible to erythromycin (MIC < 0·015–0·12 g l−1) and unable to grow on selective blood agar plates containing 2 mg l−1 erythromycin used in screening for transconjugants. The wild-type recipients were also confirmed to be susceptible to rifampicin and fusidic acid, and resistant derivatives of these, used to differentiate transconjugants from SMH036, were generated by inducing to spontaneous mutations conferring resistance to rifampicin or fusidic acid.

In vitro conjugative gene transfer of Tn1207.3

Filter mating was performed according to Giovanetti et al. (Giovanetti et al. 2002). Transconjugants were selected on blood agar plates containing either fusidic acid (25 mg l−1) or rifampicin (25 mg l−1) in addition to erythromycin (2 mg l−1). Transfers were performed in triplicate.

Confirmation of mef(A) transfer

Putative transconjugants were tested for the presence of mef(A) by PCR using primers mefF/mefR as previously described (Sutcliffe et al. 1996b). Both Tn1207.3/comEC junctions were also examined by PCR, primers TnRjF/TnRjR (D’Ercole et al. 2005) spanning the 5′ junction and comECF/orfLF (Santagati et al. 2003) spanning the 3′ junction. Donor SMH036 was used as a positive control for these PCRs. Transconjugants were randomly selected (3–6 per donor/recipient combination) for PCR validation with a total of n = 85. All recipient strains produced no amplicons in these PCRs.