Résumé
Background and objectives: Antimicrobial resistance (AMR) is a growing threat to public health, because it can compromise the treatment of certain severe infections, particularly those caused by Gram-negative bacilli such as Enterobacteriaceae, as well as invasive medical procedures and surgeries. Carbapenemase-producing bacteria (CPB) are part of the most concerning pathogen, by the production of carbapenemases, which are enzymes hydrolysing antibiotics of the carbapenems’ family. These molecules are used as a last resort, to treat severe infections [1]. Several studies have highlighted the role of hydric environment in AMR emergence, spread and persistence. Hence, antimicrobial resistant bacteria and genes should be considered as major environmental contaminant [2, 3]. However, little is known about the dynamics of CPB, particularly those producing carbapenemase in water environments. This study investigates the distribution of carbapenemase genes (blaKPC, blaOXA-48, blaNDM) and the presence of CPB in four contrasted water environments in Montpellier (France): wastewater (WW) from a university hospital and a municipal wastewater treatment plant (WWTP), and freshwater from the Lez underground spring and the Verdanson urban stream.Methods: A total of 90 water samples were collected from the four water compartments. Total environmental DNA was extracted, to quantify carbapenemase blaKPC, blaOXA-48, and blaNDM as well as the Intl-1, Intl-2 and Intl-3 genes, using qPCR. These samples were also seeded onto chromogenic media supplemented with 0.125µg/mL and 2µg/mL meropenem, allowing to constitute a collection of carbapenem-resistant bacteria, which were all identified by MALDI-TOF-MS. 160 of these strains were then screened by specific PCR to detect the three carbapenemase genes of interest.Results: The three carbepenemases genes were detected in almost all WW samples, with a maximal concentration of 4.47E+04 copies/mL in hospital WW, and 8.40E+02 in WWTP samples. However, blaNDM exhibited greater variability over time than blaKPC and blaOXA-48. The inconsistent blaNDM signal over time suggests transient contamination rather than a stable environmental reservoir. Carbapenemase genes, except blaNDM, were also detected in several freshwater samples, in up to 63% of the urban stream and in 23% of the spring water at low concentrations, reaching a maximum of 1.45 copies/mL for the urban stream, and 3.67E-02 copies/mL for the spring. In WW and urban water, blaOXA-48, blaKPC and the three integrons encoding genes follow the same temporal dynamics. Moreover, environmental factors such as precipitation and temperature had limited influence on genes load except for samples from WWTP that showed significant variations depending on temperature and seasonality (p<0.05), with a decrease of at least 1.00E+02 gene copies/mL for all the genes tested between December and February. The culture-based approach allowed the isolation of 6 864 carbapenem-resistant bacteria, mainly belonging to Pseudomonas, Acinetobacter, Citrobacter, and Enterobacter genera. Among the 160 tested strains from the hospital WW, 36 (23%) carried at least one of the three carbapenemase genes of interest, blaOXA-48 being the most frequently detected (16% of the strains). Furthermore, the majority of the carbapenemase-producing bacteria belong to environmental-associated genera Aeromonas (11/22 tested strains), Shewanella (8/15 tested strains), Citrobacter (7/14 tested strains), Comamonas (4/18 tested strains) and Pseudomonas (3/39 tested strains). Moreover, in the spring, 3/9 of the tested enterobacteria carried carbapenemase genes. Conclusions: This study provides critical elements in favor of an established environmental cycle carbapenemases genes and carbapenemase-producing enterobacteria in a french metropole: i) carbapenemases genes and carbapenemase-producing enterobacteria detection in spring groundwater (Klebsiella, Enterobacter); ii) environmental-borne carbapenemase-producing bacteria in human WW (Pseudomonas, Aeromonas, Shewanella, Comamonas...). An environmental cycle increases the risk of spreading human pathogenic bacteria carrying carbapenemase genes. Thus, there is a need for a monitoring system of environmental antimicrobial resistance. Recent developments of wastewater based epidemiology is applicable for antimicrobial monitoring, but the complexity of the WW microbial communities, which is a hot-spot for gene transfers, could blur the epidemiological signal. Monitoring antimicrobial resistance genes in fresh water could be a useful alternative to follow the epidemiology of emerging resistance with environmental cycle.