Journal of Global Antimicrobial Resistance (Dec 2022)

Genome stability during serial subculturing in hyperepidemic multidrug-resistant Klebsiella pneumoniae and Escherichia coli

  • Aline I. Moser,
  • Edgar I. Campos-Madueno,
  • Vincent Perreten,
  • Andrea Endimiani

Journal volume & issue
Vol. 31
pp. 152 – 161

Abstract

Read online

ABSTRACT: Objectives: Core-genome single nucleotide variant (cgSNV) analysis represents a powerful tool for epidemiological investigations of multidrug-resistant (MDR) bacteria. However, cgSNV thresholds to confirm whether isolates are the same clone are not formally defined. Methods: We implemented hybrid whole-genome sequencing to study the genomic changes of four MDR isolates belonging to hyperepidemic sequence types (STs) during 20 propagation steps (T20) on MacConkey and CHROMID(R) ESBL plates. The following strains were analyzed: Klebsiella pneumoniae AE-2247421 (OXA-48/NDM-1-producing, ST101), K. pneumoniae MCL-2017-2 (CTX-M-15-producing, ST307), Escherichia coli Ec-042 (OXA-181-producing, ST410), and E. coli Ec-050 (NDM-5-producing, ST167). The genome assembly at T5 and T20 was compared to that at time point zero (T0) and to two reference genomes. Results: At T20, AE-2247421 lost the IncL blaOXA-48-carrying plasmid when grown on CHROMID(R) ESBL plates, while a large fragment encompassing blaNDM-1 was lost from its IncC plasmid when grown on both plates. In contrast, no structural changes were noted for the other three strains. Regarding the cgSNVs, the following results were obtained at T5 and T20 (ranges considering the different agar plates and reference genomes): AE-2247421 (1–8 and 2–12 cgSNVs), MCL-2017-2 (both 1–2 cgSNVs), Ec-042 (both 0 cgSNVs), and Ec-050 (0–6 and 0–9 cgSNVs). Conclusion: We showed that structural changes and accumulation of cgSNVs can occur in few propagation steps under laboratory conditions. These changes might also arise in the clinical context in a short time, especially under antibiotics treatment. This phenomenon should be carefully considered because it might affect the final interpretation of epidemiological genomic analyses.

Keywords