Research ArticleHEALTH AND MEDICINE

Antibiotic resistance in European wastewater treatment plants mirrors the pattern of clinical antibiotic resistance prevalence

See allHide authors and affiliations

Science Advances  27 Mar 2019:
Vol. 5, no. 3, eaau9124
DOI: 10.1126/sciadv.aau9124
  • Fig. 1 PCoA showing the distribution of resistance and mobile genetic elements using the Bray-Curtis dissimilarity index.

    (A) Influent water from European UWTPs. The HAC and LAC countries clustered separately (R2 = 0.37, P = 0.015). The exception was for three influent samples from German UWTPs that clustered with the HAC samples (DE2 in the figure). (B) Effluent water from European UWTPs. The HAC and LAC countries did not cluster together (R2 = 0.11, P = 0.181). The significance of the cluster separation between the HAC and LAC countries was calculated using adonis with 9999 permutations. The points represent individual samples in the ordination, named according to the UWTP, and latitude is indicated by the color gradient. The point size is according to the mean total antibiotic consumption in humans from 2005 to 2015 from ECDC reports consisting of the human consumption of antibacterials for systemic use (ATC group J01) in the community (primary care sector) and the hospital sector expressed as defined daily dose (DDD) per 1000 inhabitants and per day (available at https://ecdc.europa.eu/en/antimicrobial-consumption/database/country-overview) (see also table S3). Countries: HAC—PT, Portugal; ES, Spain; CYP, Cyprus; IL, Ireland; LAC—DE, Germany; FI, Finland; NO, Norway.

  • Fig. 2 Relative gene abundance observed in influent samples from HAC and LAC countries.

    Relative abundance of (A) resistance genes and (B) mobile genetic elements. The data refer to the sum of relative abundance of amplification (ratio ARG or MGE copy number: 16S rRNA gene copy) for a given pair of primers, organized in classes of “resistance” or “transfer and recombination.” In the legend, for each gene class, the country group, HAC or LAC, with significantly higher relative abundance (P < 0.01, Mann-Whitney U test) is indicated. Samples are organized according to the sampling campaign (C2, spring 2016; C3, autumn 2016), divided by HAC and LAC. Resistance categories: AMG (aminoglycosides), MDR (multidrug resistance), SUL (sulfonamides), BL (β-lactams), MLSB, TET (tetracycline), QUI (quinolones), AMP (amphenicols), VAN (vancomycin), and others. Note: Ireland data are missing because of restrictions on influent wastewater sample collection.

  • Fig. 3 Relative gene abundance observed in effluent samples from HAC and LAC countries.

    Relative abundance of (A) resistance genes and (B) mobile genetic elements. The data refer to the sum of relative abundance of amplification (ratio ARG or MGE copy number: 16S rRNA gene copy) for a given pair of primers organized in classes of resistance or transfer and recombination. In the legend, for each gene class, the country group, HAC or LAC, with significantly higher relative abundance (P < 0.01, Mann-Whitney U test) is indicated. Samples are organized according to the sampling campaign (C1, autumn 2015; C2, spring 2016; C3, autumn 2016), divided by HAC and LAC. Resistance categories: AMG (aminoglycosides), MDR (multidrug resistance), SUL (sulfonamides), BL (β-lactams), MLSB, TET (tetracycline), QUI (quinolones), AMP (amphenicols), VAN (vancomycin), and others.

  • Fig. 4 Variation in average ARG prevalence between influent and effluent [log10(influent − effluent)] in HAC and LAC countries.

    (A) Comparison of ARG log reduction values in UWTPs of the HAC and LAC countries, where zero indicates that treatment did not affect the cumulative ARG relative abundance, positive values indicate a reduction, and negative values indicate an increase. The classes for which HAC or LAC presented higher relative abundance in the influent (P < 0.01, Mann-Whitney U test) are indicated in the table at the bottom. Asterisks indicate statistically significant differences between HAC and LAC log reduction values, and these cases are detailed in (B). (B) Comparison of average ARG relative abundance in influent and final effluent, for the classes with significantly different log reduction values in the HAC and LAC samples: (B1) MLSB, (B2) tetracyclines, (B3) insertion sequences, and (B4) transposase. The arrows indicate significant (P < 0.01; Welch’s t test) increases (↑) or decreases (↓) after treatment (C1, autumn of 2015; C2, spring of 2016; C3, autumn of 2016). WW, wastewater.

  • Fig. 5 Association between the relative abundance of ARGs in UWTP effluents (aggregated by drug classes) and country-level data on phenotypic resistance of clinical isolates [data from EARS-Net,].

    (31) Numbers represent Spearman’s rank correlations; color codes indicate statistical significance. No statistically significant correlations were observed for Acinetobacter spp., Enterococcus faecium, or S. pneumoniae.

  • Table 1 Candidate indicator assays for major gene classes in UWTP effluents.

    Assays that yielded the highest amplification in the largest number of samples are reported on the left. Assays with the most representative outcome in terms of correlation with the per-class mean are reported on the right (rho: Spearman’s rank correlation coefficient). Primer pairs corresponding to the assay are listed in table S2.

    Gene class
    (no. of assays)
    Genes dominating the classMost representative genes
    GeneAssay ID% of samplesGeneAssay IDRho
    Aminoglycoside (24)aadAAY16762.90.96
    Amphenicol (14)cmxAAY12927.1cmlAAY1270.74
    β-Lactam (61)blaOXAAY4434.50.76
    MDR (40)qacEdelta1AY15992.10.97
    MLSB (30)ermFAY2368.60.81
    Quinolone (3)qnrSrtF11AY6100.01.00
    Sulfonamide (9)sul1AY36373.00.82
    Tetracycline (34)tetQAY18559.6tetXAY1960.73
    Insertion
    sequence (9)
    ISPpsAY36986.50.98
    Integrase (7)intI1AY33680.90.96
    Transposase (10)tnpAAY20279.30.86

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/3/eaau9124/DC1

    Fig. S1. Average richness values (number of positive assays) for the different influent and effluent wastewater samples from high (HAC) and low (LAC) antibiotic consumption countries, for resistance genes and mobile genetic elements.

    Fig. S2. Food-producing animals’ antibiotics consumption [expressed in biomass (mg/kg), values for 2013] and average maximal and minimal annual temperature and precipitation (yellow, average Tmin >6°C; blue, average Tmax <5°C).

    Fig. S3. Average abundance (copies/ml; upper bars; left-hand legend) and prevalence values (gene copies/16S rRNA gene copies; lower bars; right-hand legend) for the different influent (Inf) and effluent (Eff) wastewater samples from high (HAC) and low (LAC) antibiotic consumption countries, determined on the basis of qPCR array for the genes: 16S rRNA, intI1, blaOXA, sul1, tetM, ermF, aadA, tnpA, and qacEdelta1.

    Fig. S4. Average abundance (copies/ml; upper bars; left-hand legend) and prevalence values (gene copies/16S rRNA copies; lower bars; right-hand legend) calculated by traditional real-time qPCR and qPCR array for Portuguese influent and effluent wastewater samples.

    Table S1. Influent and effluent wastewater samples used in the study.

    Table S2. qPCR primer sets and the percentage of samples that gave positive results for the influent and effluent wastewater samples.

    Table S3. Consumption of antibacterials for systemic use (ATC group J01) in the community (primary care sector) in different European countries from 2005 to 2015 (see also Fig. 1), defined as daily dose per 1000 inhabitants and per day.

    Table S4. Characterization of the UWTPs examined in this study, in terms of dimension, geographic conditions, treatment, and microbiological indicators.

    Data S1. qPCR array data.

    Data S2. List of samples.

    Data S3. List of assays.

    Reference (32)

  • Supplementary Materials

    The PDF file includes:

    • Fig. S1. Average richness values (number of positive assays) for the different influent and effluent wastewater samples from high (HAC) and low (LAC) antibiotic consumption countries, for resistance genes and mobile genetic elements.
    • Fig. S2. Food-producing animals’ antibiotics consumption expressed in biomass (mg/kg), values for 2013 and average maximal and minimal annual temperature and precipitation (yellow, average Tmin >6°C; blue, average Tmax <5°C).
    • Fig. S3. Average abundance (copies/ml; upper bars; left-hand legend) and prevalence values (gene copies/16S rRNA gene copies; lower bars; right-hand legend) for the different influent (Inf) and effluent (Eff) wastewater samples from high (HAC) and low (LAC) antibiotic consumption countries, determined on the basis of qPCR array for the genes: 16S rRNA, intI1, blaOXA, sul1, tetM, ermF, aadA, tnpA, and qacEdelta1.
    • Fig. S4. Average abundance (copies/ml; upper bars; left-hand legend) and prevalence values (gene copies/16S rRNA copies; lower bars; right-hand legend) calculated by traditional real-time qPCR and qPCR array for Portuguese influent and effluent wastewater samples.
    • Table S1. Influent and effluent wastewater samples used in the study.
    • Table S2. qPCR primer sets and the percentage of samples that gave positive results for the influent and effluent wastewater samples.
    • Table S3. Consumption of antibacterials for systemic use (ATC group J01) in the community (primary care sector) in different European countries from 2005 to 2015 (see also Fig. 1), defined as daily dose per 1000 inhabitants and per day.
    • Table S4. Characterization of the UWTPs examined in this study, in terms of dimension, geographic conditions, treatment, and microbiological indicators.
    • Reference (32)

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    Files in this Data Supplement:

Navigate This Article