Neisseria cinerea with High Ceftriaxone MIC Is a Source of Ceftriaxone and Cefixime Resistance-Mediating penA Sequences in Neisseria gonorrhoeae

ABSTRACT

Mosaic penA alleles have caused most of the cephalosporin resistance in Neisseria gonorrhoeae, but their evolution is mostly unknown. The penA gene from Neisseria cinerea strain AM1601 (ceftriaxone MIC, 1.0 μg/ml) caused ceftriaxone resistance (MIC, 1 μg/ml) in a ceftriaxone-susceptible gonococcal strain. The 3′-terminal half of AM1601 penA was almost identical to that of the ceftriaxone-resistant gonococcal GU140106 and FC428 strains. N. cinerea can serve as a reservoir of ceftriaxone resistance-mediating penA sequences that can be transferred to gonococci.

TEXT

Resistance to the extended-spectrum cephalosporin (ESC) ceftriaxone in Neisseria gonorrhoeae has sporadically emerged worldwide. The main ESC resistance determinant is mosaic penA alleles, encoding penicillin-binding protein 2 (PBP2) (1–10). These gonococcal mosaic penA alleles are proposed to have evolved through transformation of partial penA sequences from commensal Neisseria species (1, 2, 8, 11, 12); however, detailed knowledge is lacking. The ceftriaxone-resistant gonococcal strains H041 (4), A8806 (6), GU140106 (7), and FC428 (8) possessed different mosaic penA alleles. However, A8806 (6), GU140106 (7), and FC428 (8) possessed identical or almost identical 3′-terminal halves of penA, although the central region of penA showed substantially less nucleotide sequence similarity (8). The conserved 3′-terminal part of penA_FC428 was very different from the FA1090 wild-type penA (8), illustrating that the origin of gonococcal ceftriaxone resistance exists in other species.

Here, we investigated Neisseria cinerea AM1601, which was isolated in 2016 from a patient with bacteremia in Aichi, Japan (13), and has high MICs for ceftriaxone (1.0 μg/ml) and cefixime (2.0 μg/ml), as determined by an agar dilution method according to CLSI guidelines (14). To verify the species of N. cinerea AM1601, ribosomal multilocus sequence typing was performed. Briefly, the single-nucleotide polymorphisms (SNPs) on 53 rps alleles were extracted from the AM1601 genome sequence generated by a MiSeq sequencer (Illumina, San Diego, CA, USA) and were compared with those on 53 rps alleles from various Neisseria species (15). The results indicated that N. cinerea AM1601 clustered with six N. cinerea reference strains (see Fig. S1 in the supplemental material) (15, 16).

To show that penA of N. cinerea AM1601 is responsible for ceftriaxone resistance, a full-length AM1601 penA gene was PCR amplified using PrimeStar HS Premix (TaKaRa Bio, Shiga, Japan), genomic DNA, and the primers 5′-ATGTTGATTAAGAGCGAATATAAG-3′ and 5′-TTAAGACGGTGTTTTGACGG-3′. The primers were designed on the basis of the N. cinerea AM1601 penA sequence determined by whole-genome sequencing and confirmed by conventional Sanger sequencing, as described previously (8). After purification with the High Pure PCR product purification kit (Roche Diagnostics GmbH, Mannheim, Germany), the PCR product was transformed into NG9807, as described previously (17). The ceftriaxone-susceptible gonococcal strain NG9807 (MIC of 0.016 μg/ml, with the wild-type penA) was used as a recipient for transformation (17). NG9807 has a single-nucleotide (A) deletion in the inverted repeat of the mtrR promoter, a PBP1 L421P alteration, and penB alterations at PorB1b G120 and A121 (4). The transformation frequency was estimated as 1 in 10⁷ recipient cells. The full-length penA gene in the NG9807 transformant was sequenced using conventional Sanger sequencing.

Transformation of penA_AM1601 into the ceftriaxone-susceptible gonococcal strain NG9807 caused resistance to ceftriaxone (64-fold MIC increase, from 0.016 to 1 μg/ml) and cefixime (128-fold MIC increase, from 0.016 to 2 μg/ml) (14, 18). Only ∼62% of the donor penA_AM1601 was incorporated into the recipient penA_NG9807 (Fig. 1), resulting in a mosaic penA allele. To characterize the penA sequence of N. cinerea AM1601, we compared the penA gene of AM1601 with the penA genes of ceftriaxone-resistant N. gonorrhoeae strains using ClustalW. The levels of similarity between N. cinerea penA_AM1601 and ceftriaxone-resistant gonococcal penA_H041 (4), penA_GU140106 (7), and penA_FC428 (8) were 92.7%, 90.3%, and 92.9%, respectively. The mismatches between penA_AM1601 and penA_GU140106 accumulated in the 5′-terminal half, whereas there were no SNPs in the 3′-terminal half, including the PBP2 β-lactam-active motifs (Fig. 2A). Similarly, the penA_AM1601 3′-terminal half was highly homologous to the corresponding penA_FC428 sequence (only one synonymous SNP) (Fig. 2B). Accordingly, the AM1601 PBP2 shared a trait with those of the ceftriaxone-resistant gonococcal strains GU140106 (7) and FC428 (8). Importantly, the PBP2 forms of all of these strains possessed V311 and S483, which are two of the three mutated amino acids causing high-level ceftriaxone resistance in N. gonorrhoeae H041 (19). However, the penA 3′-terminal region that was conserved between penA_AM1601 and penA_H041 was smaller (Fig. 2). To further characterize penA_AM1601, it was compared with another penA (GenBank accession number AB904039), from a Neisseria strain that has high ESC MICs (ceftriaxone MIC, 2 μg/ml; cefixime MIC, 4 μg/ml). The strain was verified as N. cinerea by rps gene comparison in this study (Fig. S1). N. cinerea SH43-3 was isolated in 2013 from an asymptomatic female sex worker, in Kyoto, Japan, during a routine examination for sexually transmitted infections (gonococcus-negative pharyngeal specimen). The 5′-terminal and 3′-terminal parts were very similar (99.5%) and identical (100%), respectively. Nearly all of the mismatches accumulated in the central part of penA, indicating recombination event(s) (Fig. 2D).

FIG 1.

Mosaic penA allele generated by transformation of Neisseria cinerea penA_AM1601 into Neisseria gonorrhoeae NG9807 (with wild-type penA). The penA allele of the NG9807 transformant (penA_TF) was compared with the donor allele of N. cinerea AM1601 (penA_AM1601) and the recipient wild-type allele of N. gonorrhoeae NG9807 (penA_NG9807). (A) Number of mismatches in each 50 bp between alleles penA_TF and penA_AM1601. (B) Number of mismatches in each 50 bp between alleles penA_TF and penA_NG9807. (C) Schematic representation of the mosaic penA_TF by a white rectangle (remaining from penA_NG9807) and a black rectangle (derived from penA_AM1601).

FIG 2.

Sequence comparison of penA genes from ceftriaxone-resistant Neisseria cinerea and Neisseria gonorrhoeae strains isolated in Japan. Pairwise comparisons of penA nucleotide sequences for penA genes from N. cinerea AM1601 and N. gonorrhoeae GU140106 (7) (A), N. gonorrhoeae FC428 (8) (B), N. gonorrhoeae H041 (4) (C), and N. cinerea SH43-3 (D) are shown. The mismatched bases in each 50 bp of the penA genes were counted. The similarity of each region is indicated, with the nucleotide positions evaluated in parentheses. *, one synonymous mismatch, at nucleotide position 1296; **, three mismatches, at nucleotide positions 48, 219, and 489.

Commensal Neisseria species, including N. cinerea, are members of the human oropharyngeal microflora (20–23). These Neisseria species might be genetic reservoirs of resistance determinants for β-lactam antimicrobials (including ESCs) that can be transferred to the pathogenic species Neisseria meningitidis and gonococci (1, 11, 12, 23–26). We demonstrate that N. cinerea strains with high ceftriaxone MICs (1 to 2 μg/ml) possess ceftriaxone resistance-mediating penA sequences that can be transferred to gonococci by transformation and result in ceftriaxone and cefixime resistance. The 3′-terminal half of mosaic penA in the transformant, which was transferred from N. cinerea, has also been described in ceftriaxone-resistant clinical gonococcal strains, i.e., A8806 (6), GU140106 (7), and FC428 (8), isolated in 2013 to 2015 in Australia and Japan. Accordingly, the 3′-terminal half of N. cinerea penA has caused ceftriaxone resistance in genetically different gonococcal strains in different countries. This indicates that N. cinerea strains represent an origin of the ceftriaxone resistance-mediating penA sequences in the gonococcal strains A8806 (6), GU140106 (7), and FC428 (8). However, an unknown origin of ceftriaxone resistance-mediating penA sequences might also exist, from which genetic material has been transferred to both ceftriaxone-resistant gonococcal strains and ceftriaxone-resistant N. cinerea strains, which might be supported by the mismatches in the central region of penA in AM1601 versus SH43-3. Further investigations of commensal Neisseria species are imperative.

In conclusion, N. cinerea can serve as a reservoir of ceftriaxone resistance-mediating penA sequences that are transferred to and cause ceftriaxone and cefixime resistance in clinical gonococcal strains. Examinations of commensal Neisseria species are crucial to understand, and ideally to mitigate, the emergence and evolution of resistance to ESCs in gonococci. This will provide new insights regarding interspecies sharing and reservoirs of resistance determinants for other antimicrobials in commensal bacteria.

Accession number(s).

The N. cinerea AM1601 penA sequence was deposited in the DDBJ (accession number LC316656).