Increasing rates of erm(B) and erm(N) in human Campylobacter coli and Campylobacter jejuni erythromycin-resistant isolates between 2018 and 2023 in France

Abstract

Macrolides are the first-line compounds used for the treatment of campylobacteriosis. Macrolide resistance remains low in France, with mutations in 23S rDNA being the main associated resistance mechanism. However, two erythromycin methyltransferases have also been identified: erm(B), which is mainly described in animal reservoirs, and erm(N), which is strictly described in humans. In France, between 2018 and 2023, erythromycin-resistant Campylobacter species strains were systematically sequenced and analyzed via an in-house bioinformatics pipeline, leading to the identification of the resistomes, MLST and cgMLST, as well as the characterization of the source of contamination. In this study, the genomes of 280 erythromycin-resistant strains were sequenced over a 6-year period. The identification of erythromycin-associated resistance markers revealed a predominance of 23S rDNA mutations, in 90% of cases, but also erm-type methyltransferases in 10% of cases: 75% for erm(N) and 25% for erm(B). Over this period, an important increase in the rate of erm-positive isolates was observed: 2% in 2018 compared with 13% in 2023, with 10% for erm(N) and 3% for erm(B). erm(N) has been found exclusively within a CRISPR-Cas9 operon, whereas erm(B) has been found within diverse types of resistance genomic islands. Each erm(N)- or erm(B)-positive isolate had at least two other resistance markers (mostly ciprofloxacin, tetracycline, or ampicillin) and often carried aminoglycoside-associated resistance genes. The majority of the erm-positive isolates were obtained from chicken. The increasing rates of erm-positive and multiresistant isolates make the monitoring of erythromycin-resistant Campylobacter strains, specifically within the chicken meat production, a topic of serious importance.

Keywords: Campylobacter; NGS; macrolide; methyltransferase; resistance.

Fig 1

Core-genome MLST tree of all 280 studied C. coli and C. jejuni clinical isolates. Core-genome profiles were identified via the Campylobacter scheme v2.0 from PubMLST, and the tree was displayed using MEGA software combined with the iTOL online tool. C. coli isolates are highlighted in orange, whereas C. jejuni isolates are highlighted in green. Various STs and CCs were found, and their combinations were attributed to a specific color. Furthermore, “ST-?” or “CC-?” annotations were used to display undefined STs or undefined CCs, respectively, and “<1%” annotation was used to display STs or CCs with fewer than 1% of the studied isolates, being unique STs or CCs identified in this study. Dots on branches indicate a bootstrap score of 100%.

Fig 2

Evolution of C. jejuni and C. coli erythromycin-resistant clinical isolates between 2018 and 2023 in France, with the associated resistance mechanism proportions. The left y-axis displays erythromycin resistance rates in France between 2018 and 2023; orange represents C. coli clinical isolates (n = 1,077 isolates tested per year on average, data not included), and green represents C. jejuni clinical isolates (n = 6,870 isolates tested per year on average, data not included). These data are based on NRCCH annual reports (www.cnrch.fr/). The right y-axis shown with stacked bars indicates the proportion of each resistance mechanism associated to erythromycin for each year and the isolates that were sequenced in the present study. The total number of erythromycin-resistant isolates per year is indicated above the corresponding stacked bar.

Fig 3

Erythromycin minimum inhibitory concentration distributions from the agar dilution method. Boxplots were drawn using GraphPad Prism from MICs (mg/L) from a selection of all erm-positive isolates (erm(N): n = 7; erm(B): n = 21), and 33 23S rDNA-mutated isolates (A2074T: n = 9; A2074C: n = 8; A2074G: n = 7; A2075G: n = 9). A nonparametric Mann–Whitney test revealed a significant difference between 23S-mutated isolates and erm-positive isolates (23S vs. erm(N): P < 0.001**; 23S vs. erm(B): P = 0.65 ns; erm(N) vs. erm(B): P = 0.008**).

Fig 4

CRISPR–Cas9 operons of each erm(N)-positive C. coli clinical isolate. CRISPR–Cas9 regions were extracted from assembly data between the cas9 and moaE genes. Various types of CRISPR arrays were identified and are indicated in colored boxes as follows: type I in red, II in blue, III in orange, IV in green, and V in purple. The number within each type indicates exogenous sequences, and “:” indicates the C. coli palindromic repeat sequence “ATTTTACCATAAAGAAATTTAAAAAGGGACTAAAA.” The exogenous sequences are as follows: 1 = CCTATTGCAACCCTTGTTTCACGACTATAA; 2 = TTTGCAAGATAGTGATTTAAGAGATGCTTT; 3 = AAGTTTTGAAACAAGAGTGTATTATGATTA; 4 = CACCCTTCCAAAAGGGTGGAGAAGGGTTTA; 5 = GTTTTTATTTGTGGTTATAAAATAAAAAAG; 6 = TTCATAGCATCTTGCGAGCTTTTAAAGGCA; 7 = TTGCAAGATAGTGATTTAAGAGATGCTTT. The sequences for cas9, cas1, cas2, erm(N), and moeA, as indicated by percentages in the figure, were almost identical among all the isolates. The erythromycin MICs highlighted here are those obtained via the agar dilution method. The isolate “ISO1-2016”, not included in the present study, is used here as an example of a type I erm(N) isolate, as previously described (12).

Fig 5

Chromosomal multidrug resistance genomic islands (MDRGI) of each erm(B)-positive isolate. The MDRGI was extracted from the assembly data at an average of −6/+6 genes surrounding erm(B) (in red). Genes annotated as tet(O) using Prokka are displayed in purple, and other resistance genes are in yellow. The remaining genes are not related to AMR or correspond to hypothetical genes (Hp = hypothetical protein; hem = bacteriohemerythrin; php = phosphorylase; tam = trans-aconitate 2-methyltransferase; IS1216E = transposase). MDRGI types were defined based on the alignment of raw sequencing data against 11 types defined in previous publications (33, 34). Undefined types are indicated as “?” red boxes. The erythromycin MICs highlighted here are those obtained via the agar dilution method.