Frequent Tn2 Misannotation in the Genetic Background of rmtB

LETTER

RmtB is a 16S rRNA methyltransferase that confers a high level of resistance to both old and new aminoglycoside agents (1). rmtB was first reported in a Serratia marcescens clinical isolate from Japan in 2002 (2). Since then, it has been widely reported in isolates recovered from environmental and hospital settings. Usually, IncFII plasmids harboring rmtB also carry carbapenemase-encoding genes that make its spread easier (3, 4). The rmtB genetic background is frequently related to bla_TEM-1, which is usually located upstream of rmtB and is recognized as a Tn3 transposon passenger gene (tnpA, tnpR, and bla_TEM-1) (2–6). However, this β-lactamase-encoding gene has been encountered in diverse transposons, like Tn1, Tn2, and Tn3, which were the first transposons characterized as related to antimicrobial resistance. Although Tn2 and Tn3 show high similarity, sharing the same mechanisms of transposition, they possess distinct passenger genes, like bla_TEM-1a in Tn3 and bla_TEM-1b in Tn2 (7). In addition, other nucleotide differences can be found in transposase- and resolvase-encoding genes and at the resolution site (res). The main mutations are usually found in a region located between nucleotides 94 and 150 at the res site (7). In this manner, the four Tn2 sequences regions (tnpA, res site, tnpR, and bla_TEM) contained genetic characteristics capable of differentiating the other related transposons. Based on that, we analyzed the sequences of 19 whole-plasmid sequences carrying rmtB. We also studied 15 partial sequences of rmtB-associated transposons, which were deposited in GenBank to investigate the rmtB genetic background and its association with the bla_TEM-1 variants.

Three conserved rmtB genetic backgrounds (tnpA, tnpR, bla_TEM-1b, and rmtB or tnpA, ΔtnpR, bla_TEM-1b, and rmtB or ΔtnpA, tnpR, bla_TEM-1b, and rmtB) were observed in all 34 analyzed sequences (Fig. 1A). In all structures, bla_TEM-1b was detected. In this manner, we conclude that rmtB is in fact related to Tn2 and not Tn3, as suggested by many authors (2, 3, 5, 6). bla_TEM-1a and bla_TEM-1b show three nucleotide differences (₁₈C→T, ₂₂₈C→T, ₃₉₆G→T) but encode identical TEM-1 enzymes. Although silent, these mutations may be helpful in identifying and suggesting the evolution of mobile genetic elements like Tn2 and Tn3. According to Partridge and Hall (7), the similarity between Tn2 and Tn3 has corroborated the misannotation of Tn2 as Tn3. In fact, a Tn2 or ΔTn2 transposon located upstream of rmtB has been recognized by only a few authors (8–12). Two factors have contributed to the misannotation of Tn2 as Tn3: the high similarity found between these two transposons and an annotation error in the databases. When a Tn2 nucleotide BLAST analysis is performed using the ISfinder, Tn3 is the first obtained hit (E value, 0.0; score, 8,387; 97% identity). UniProt and NCBI BLAST analyses also indicate the Tn3 transposon as the main result, with rare entries for Tn2. To make easier the proper automatic annotation, the Tn2 sequence was submitted to the ISfinder database. We also confirmed the presence of Tn2 by in silico MslI digestion using SnapGene version 3.3 (8). This analysis was able to differentiate Tn2 from its correlated variant Tn2a (Fig. 1B). Tn2a was described in 2011 (8) and differs from Tn2 by 9 nucleotides located in the resolution site.

• Open in new tab
• Download powerpoint

FIG 1

(A) Comparison of the main transposon sequence regions (tnpA, res, tnpR, and bla_TEM) with Tn3. Genes are shown by labeled arrows, inverted repeats of Tn3 are shown as black triangles, and the resolution site is a dark-gray box. Inverted repeats are located in the same positions of Tn2. Vertical black lines indicate differences in the nucleotide sequences of Tn3 (reference sequence) (GenBank accession number HM749966 ) from those of the other analyzed transposons. The red box in pCT-KPC and the p477Kp transposons indicate the presence of truncated tnpR. Insertion sequences were identified by ISfinder and are demonstrated by light-purple arrows. The pEC012 plasmid shows an insertion of a partial IS26 (ΔIS26) sequence and an IS15DI sequence that show 99% similarity with IS26, according to ISfinder. Purple arrows indicate the presence of rmtB downstream of Tn2 transposons in pHeBE7 (Tn2), pCT-KPC (ΔTn2), p477Kp (ΔTn2), pCC1409-1 (ΔTn2), and pEC012 (Tn2). This figure was compiled from alignments created using SeaView version 4.6.1. (B) In silico MslI restriction analysis of Tn2 and its variants proposed by Bailey and colleagues (8). Lane 1, Tn2 of pHeBE7; lane 2, Tn2 of pIP1206; lane 3, Tn2 of pMC-NDM; lane 4, ΔTn2 (tnpR and bla_TEM-1b) of pC06114; lane 5, ΔTn2 of p397Kp (ΔtnpR and bla_TEM-1b). In both ΔTn2 sequences analyzed, tnpA was missing. MslI restriction confirmed the presence of Tn2 in isolates with the complete transposon present (tnpA, tnpR, and bla_TEM-1b). However, this methodology was shown to not be accurate in identifying the ΔTn2 sequence, since the digested fragment comprises the end of tnpA (absent in ΔTn2) and the beginning of bla_TEM-1b. MW, molecular weight markers.

The origin of the rmtB and Tn2 association remains unclear, since the right and left inverted repeats (IRR and IRL, respectively) of Tn2 have been located downstream of tnpA and upstream of bla_TEM-1b (Fig. 1A), respectively. The absence of these sequences upstream of rmtB suggests the nonmobilization of rmtB by Tn2. For this mobilization to occur, the IRL upstream bla_TEM-1b gene would have been inactivated by a new passenger gene (in this case, rmtB) and a more distant “surrogate” IR sequence recruited (13). This kind of mobilization had occurred with Tn4401, a derivate transposon of Tn3 that carries bla_KPC-2 (14). Despite this genetic evidence, transposition in vitro experiments are necessary to confirm this hypothesis.

In conclusion, we reinforce the need for a proper annotation of the rmtB genetic structure through manual checking of automatic annotation to avoid a misperception about an epidemiological change. To date, rmtB has always been associated with Tn2 and never with Tn3.