Unexpected relationships between frequency of antimicrobial resistance, disease phenotype and emm type in group A Streptococcus

Abstract

Despite universal susceptibility to β-lactams, resistance to second-line antimicrobials (e.g. erythromycin) is increasingly common among group A Streptococcus (GAS). To better understand the frequency of regional GAS antimicrobial resistance, we screened a previously described GAS strain collection from Houston, TX, USA, for resistance to commonly used antimicrobials. A total of 100/929 (10.8 %) showed resistance to at least one antimicrobial. Tetracycline resistance was identified in 52 (5.6 %) GAS strains. The cumulative frequency of erythromycin and clindamycin resistance [macrolide (M) and macrolide-lincosamide-streptogramin (MLS) phenotypes] was greatest among invasive GAS strains (9.9 %) compared to that of strains derived from any other infection type (5.9 %, P=0.045). We identified emm types 11, 75, 77 and 92 as the only emm types with high (e.g. >50 %) within-emm type resistance and contributing to the majority (24/26; 92 %) of erythromycin/clindamycin resistance in invasive GAS. High-frequency resistance emm types were also significantly overrepresented in invasive GAS strains as indicated by invasive index. We performed whole-genome sequencing to define genetic elements associated with resistance among emm types 11, 75, 77 and 92. Diverse mobile elements contributed to GAS resistance including transposons, integrative conjugative elements, prophage and a plasmid. Phylogenetic analysis suggests recent clonal emergence of emm92 GAS strains. Our findings indicate that less frequently encountered GAS emm types disproportionately contribute to resistance phenotypes, are defined by diverse mobile genetic elements and may favour invasive disease.

Keywords: antimicrobial resistance; emm type; group A Streptococcus; invasive; pharyngitis; skin and soft tissue.

Fig. 1.

Frequency of antibiotic resistance by emm type and association with invasive disease in GAS strains from Houston, TX, USA. (a) Total number of isolates by emm type (left y-axis) and frequency of any antibiotic resistance (red bars). Only those emm types with at least 10 observations are indicated. Cumulative frequency of resistance is indicated by the red line (right y-axis). (b) Invasive index versus frequency of antibiotic resistance by emm type in the Houston GAS strains. Cumulative invasive indices and resistance ratios (see Methods and Table S1) are indicated between low-frequency (blue shaded circle) and high-frequency (red shaded circle) resistant emm types (Fisher’s exact test). emm92 GAS strains are not shown as no pharyngeal isolates were identified in the Houston GAS collection for calculation of the invasive index.

Fig. 2.

blast comparisons of mobile elements harbouring antibiotic-resistance genes identified among emm11, emm75, emm77 and emm92 GAS in Houston, TX. (a) Tn916 and Tn916-like (Tn6002 and Tn6003) elements identified in emm11, emm75 and emm77 GAS strains. Tn6003 contains two copies of erm(B) (red) that may facilitate excision of the macrolide-aminoglycoside-streptothricin (MAS) element in formation of Tn6002 [14]. (b) M75.2 prophage containing Tn1207.1 (red shaded box) compared to the previously described ϕ1207.3 [15]. (c) ICE identified in emm77 GAS strains. ICESpyM77.2 is identical to that in NCTC13742 (NCBI accession number LS483386) and differs from ICESpyM77.1 by the absence of a 16.6 kb segment containing erm(A) (red). (d) ICE identified in emm92 (ICESpyM92) compared to ICE in NGAS322 (ICESpyM114) [18]. Shaded regions represent >95 % nucleotide identity in all comparisons.

Fig. 3.

Phylogenetic reconstruction using 827 core biallelic SNP loci of emm92 GAS strains from Houston (n=19) and the CDC ABCs (n=239) relative to the reference TSPY556. Node colour and shape indicate year of isolation and geographical location, respectively (see node legend). Resistance genotypes for aminoglycosides (AG), MLS and tetracycline (Tet) are shown on right, and coloured by allele (see resistance legend).