LAUSR

We previously described two identical extensively drug-resistant (XDR) Haemophilus parainfluenzae isolates non-susceptible to β-lactams, macrolides, quinolones, tetracycline and chloramphenicol [1]. In an effort to analyse how the resistance genes of these pathogens were organised, one isolate (AE-2096513) underwent Illumina HiSeq 200 whole-genome sequencing. Genomic DNA was extracted and multiplexed paired-end libraries were created using the sample preparation kit from Illumina Inc. (San Diego, CA). SPAdes v.2.5 (kmer sizes = 33, 55, 67, 81, 91, 93, 95, 97, 99) was used for read correction and de novo assembly [2]. The resulting number of scaffolds and total sequence length was 2.71 Mb (GenBank accession no. MOTP00000000). Scaffolds with a length of <500 bp were extracted, resulting in 124 scaffolds that had more than 100-fold read coverage. Special emphasis was then placed on the genetic environment of mef(E)/mel (encoding a dual efflux pump affecting macrolides), often found within the macrolide efflux genetic assembly (MEGA) element, and tet(M) (encoding a ribosomal protective protein against tetracyclines). The Illumina reads weremapped using Bowtie2 v.2.1.0 to the MEGA element (accession no. FR671415). The reads pairs, of which only one mate mapped to the MEGA element, were extracted. The non-mapping mates were aligned to the genome sequence of the reference H. parainfluenzae T3T1 (accession no. YP_004823607) using Basic Local Alignment Search Tool (BLAST) to specifically focus on the flanking regions of the tet(M)–MEGA element (accession no. KJ545575; see results below). Using this information, a fusion sequence including the tet(M)– MEGA element was constructed in silico (Fig. 1). Based on this result, sequence primers were designed to confirm the chromosomal insertion site of the element using traditional Sanger sequencing. First, using the tet(M)–MEGA flanking primers RF-end (5′GGTTTAACCGAATGGGCGCCTGC-3′) and 19210rev (5′-CTTGCAGT CAAATAGTAGTTGG-3′), a PCRproductwith theexpected insert site (ca. 10 kb) was obtained. Second, it was revealed that the tet(M)–MEGA elementwas inserted 2 bpdownstreamof RF-3 (ORF19190) at position 1 989 391 compared with H. parainfluenzae T3T1 using primers RFendandMEGA-end(5′-TTTATTTAAGAATACCTTGCCGC-3′).Third,primers mef(E)F (5′-TTCTTCTGGTACTAAAAGTGG-3′) and 19210rev were used to obtain a PCR product of 2.1 kb, which was subsequently sequenced using an additional internal primer (IMLS2, 5′-TCCCGCACCA TTTATCAGCTT-3′). This revealed that the downstream insertion site of tet(M)–MEGA in AE-2096513 is at position 1 989 935 compared with T3T1, therefore replacing ORF19200 (Fig. 1). Finally, we closely inspected the ends of the element, showing that there are no inverted repeats rendering difficult a possible explanation for the mechanism of insertion (Fig. 1). The MEGA element has been previously reported in streptococcal species such as Streptococcus salivarius and Streptococcus pneumoniae [3]. For S. salivarius, no adjacent tet(M) was described, whilst for S. pneumoniae tet(M) upstream of the MEGA element and within transposon Tn916 has been frequently identified (Fig. 1). Tn916 and the often co-identified composite transposons (e.g. Tn2009 and Tn2010) are thought of being mobile by transformation, and more than four distinct insertion sites have been identified within the pneumococcal genome [4]. In AE-2096513, tet(M) together with MEGA were in a ‘non-Tn916’ context because Tn916 typically contains the ORF8 to 23 (as illustrated for S. pneumoniae 23771) that were absent in our element (Fig. 1). In conclusion, we identified for the first time the tet(M)–MEGA element in an H. parainfluenzae isolate. This element is probably mobilised and might be transferred between different members of the microbial community. In the case of H. parainfluenzae AE2096513, this strain has been isolated along with a pan-susceptible Neisseria gonorrhoeae from a urethral swab [1]. H. parainfluenzaewas therefore considered as a coloniser, but it would be particularly worrisome if the newly described tet(M)–MEGA could be exchangedwith N. gonorrhoeae, conferring resistance to azithromycin, one of the standard therapeutic options [5]. However, H. parainfluenzae is normally isolated from the pharynx, where there are also streptococci and pneumococci. Considering the close similarity of the tet(M)– MEGA elements of H. parainfluenzae AE-2096513 and S. pneumoniae 23711, genetic exchange between these two bacterial species is likely. These findings may therefore serve as a paradigmatic example of microbial acquisition of mobile genetic elements carrying resistance genes against clinically used antibiotics.