Evolution of IS26-bounded pseudo-compound transposons carrying the tet(C) tetracycline resistance determinant

Highlights

• A large FII-18:FIB-1 plasmid, pCERC14, from a commensal E. coli was sequenced.

• pCERC14 includes genes for iron uptake systems and for colicin M synthesis.

• dfrA5-sul1 and tet(C) are in 2 IS26-bounded pseudo-compound transposons (PCT).

• A potential progenitor for the PCT, PTntet(C)-var, was identified.

• IS26-mediated deletions have generated some widely-distributed PTntet(C) variants.

Abstract

A large plasmid, pCERC14, found in an antibiotic resistant commensal Escherichia coli isolate recovered from a healthy adult was sequenced. pCERC14 was 162,926 bp and carried FII-18 and FIB-1 replicons and an F-like transfer region as well as several virulence determinants, some of which are involved in the uptake of iron which would be advantageous for the commensal lifestyle. The plasmid backbone is interrupted in 11 places by complete IS (IS1 (4 copies), IS2 (2), IS629 (2) and single IS100, IS186, ISEc33) and in three places by partial IS copies. The antibiotic resistance genes were found in two IS26-bounded pseudo-compound transposons (PCT). One contained a remnant of a class 1 integron that includes a dfrA5 gene cassette and the sul1 gene conferring resistance to trimethoprim and sulphonamides, respectively. The second, named PTntet(C)-var, contained a 4828 bp DNA segment that includes the tet(C) tetracycline resistance determinant. As tet(C) is relatively rare in E. coli and other Gram-negative bacterial isolates, the structure and evolution of tet(C)-containing PCT in available sequences was examined. The largest identified was PTntet(C), a close relative of PTntet(C)-var, and a potential progenitor for these PCT. Most PCT shared one internal boundary with PTntet(C) but the length of the central tet(C)-containing segment was shorter due to IS26-mediated deletions. The most abundant variant form, previously named Tn6309, was widely distributed and, in a derivative of it, most of the tetA(C) gene has been replaced by the tetA(A) gene presumably by homologous recombination.

Introduction

In Gram-negative bacteria, insertion sequence IS26 is often associated with the formation of structures that are bounded by copies of this IS and resemble compound or composite transposons, e.g. (Wrighton and Strike, 1987; Schnabel and Jones, 1999; Cain and Hall, 2012; Blackwell et al., 2017) and, as a consequence, it is a major player in the dissemination of a variety of different antibiotic resistance genes (Partridge et al., 2018). However, it has been long known that structures with inversely-oriented IS26 cannot relocate and those with directly-oriented IS26 copies are not strictly compound transposons as they are unable to move as a single discrete unit (Galas and Chandler, 1989; Harmer et al., 2020). Rather, a step catalysed by the IS26 transposase and a homologous recombination step are needed (see Harmer et al. (2020) for details). Briefly, if one of the surrounding IS26 copies causes formation of a cointegrate with a target DNA molecule, the cointegrate structure formed can only be resolved by homologous recombination between two of the three directly-oriented IS26 copies in the cointegrate. Recombination between one of the three possible pairs results in transfer of the transposon to the target molecule (Brown et al., 1984; Nies et al., 1985; Galas and Chandler, 1989; Mahillon and Chandler, 1998). Alternatively, homologous recombination can generate a circular translocatable unit (TU) which can be incorporated elsewhere. Hence, structures bounded by directly-oriented IS26 copies have been termed a “pseudo-compound transposon” as their movement differed from transposition of classic compound transposons (Galas and Chandler, 1989; Harmer et al., 2020). Here, we abbreviate “pseudo-compound transposon” to “PCT” and specific structures are referred to as PTn for pseudo transposon. It is well known that IS26 can cause deletions of adjacent DNA (Hall, 1987; Partridge et al., 2011; Blackwell et al., 2015; Nigro and Hall, 2016) and this can potentially lead to differences in the length of the resistance gene-containing DNA fragment located between the IS leading to a series of related IS26-bounded PCTs.

Over thirty classes of tetracycline resistance determinants have been described (Levy et al., 1999; Roberts and Schwarz, 2016). However, relatively few classes are found widely in Gram-negative bacteria. In these organisms, five classes of tetracycline resistance determinants, tet(A), tet(B), tet(C), tet(D) and tet(G) each confer resistance by encoding an energy-dependent efflux protein TetA, and a TetR transcriptional repressor of the tetA gene (Roberts, 2005). The tet(A) and tet(B) determinants are more common than tet(C) and tet(D) (Bailey et al., 2010; Koo and Woo, 2011; Anantham and Hall, 2012). The tet(C) determinant has previously been associated with IS26-bounded transposon-like structures or PCTs (Schnabel and Jones, 1999; Stokes et al., 2007; Evershed et al., 2009; Turner et al., 2014; Anantham et al., 2015; Sun et al., 2016), as has tet(D) (Anantham et al., 2015).

A single isolate in our collection of over 50 antibiotic resistant commensal E. coli recovered from healthy adults was found to carry a tet(C) determinant (Anantham and Hall, 2012). This isolate, 2.3-R4, is resistant to tetracycline and also resistant to sulfamethoxazole due to the presence of the sul1 gene and to trimethoprim (dfrA5 gene) (Anantham and Hall, 2012; Moran et al., 2017). It was later shown by PCR-based typing to harbour FII and FIB plasmid replicons consistent with the presence of an FII:FIB plasmid (Moran et al., 2015). Here, we report the complete sequence of pCERC14, an FII:FIB plasmid found in 2.3-R4 which carries a tet(C) determinant. The relationships between the IS26-bounded tet(C)-containing PCTs found in pCERC14 and other tet(C)-carrying PCT sequences found in GenBank was investigated.