The quality criteria used to select the qPCR array data to be used in the analyses were (i) samples with at least two replicates, (ii) samples with more than 5 ng of DNA, and (iii) quantifications with a Ct ≤ 27. A Ct value of 27 corresponded to a limit of quantification lower than 100 gene copies (25). The qPCR array semiquantitative analyses did not support estimations based on gene abundance per volume of water because calibration curves were not obtained for each gene. However, for a subset of 10 genes, calibration curves that supported the assessment of the relationship between relative (16S rRNA based) and absolute quantification (fig. S3) were obtained. Some samples (Portugal) used for the array were also tested by real-time qPCR for the quantification of selected genes (fig. S4). Those analyses allowed us to conclude that the overall trends are similar between the quantitative qPCR and the semiquantitative qPCR array.

The 384 primer sets target 259 different genetic determinants. These comprised 229 ARGs organized in classes according to the class of antibiotics to which they confer resistance (AMG, aminoglycosides; SUL, sulfonamides; BL, β-lactams; MLSB; TET, tetracycline; QUI, quinolones; AMP, amphenicols; VAN, vancomycin; and others), as MDR (multidrug resistance) when conferring resistance to more than one class of antibiotics; 16 genetic transfer and recombination elements (integrase, transposase, and insertion sequence); 9 plasmid-associated genetic determinants; and 5 housekeeping genes. The plasmid-associated assays were excluded from the analyses because of the poor coverage (table S2). All the analyses were performed considering the remaining 375 primer pairs independently. The results of amplicons’ relative abundance were calculated using the Ct values of the reference gene and the genes of interest, applying the following formula: (2(Ctreference geneCttarget gene)) (26).

The statistical analyses were performed using the R environment v3.4.0. To compare the average values between two groups, Mann-Whitney U tests (27) were performed. Correlations between the relative abundance of each gene and the sum of the relative abundance of all genes from the same group (Table 1) were computed using Spearman’s rank-based approach (28). Statistical modeling of the total abundance of ARGs involved a power transformation of the gene’s relative abundances and the complete removal of zero-inflated assays (no measurable amplification in >25% of the samples).

All reported P values were adjusted for multiple testing according to the respective hypotheses (29). To calculate the log reduction of the abundance of resistance and genetic transfer and recombination classes, the formula log10(influent relative abundance) − log10(effluent relative abundance) was applied, and Mann-Whitney U test was used to find significant differences between HAC and LAC. PCoA was calculated using the Bray-Curtis dissimilarities calculated using the package vegan v2.4 (30) and the cmdscale command from stats package. The latitude was defined by the latitude of the city of the UWTP. The data for human consumption were collected from the antimicrobial consumption database (available at, considering the 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 using the mean from years 2005–2015.

To study the association between relative ARG abundance in UWTP effluents and phenotypic resistance in the primary care sector [data from EARS-Net, (31); Fig. 5], the information from the two sources was matched in terms of target antibiotics. Specifically, phenotypic resistance against penicillins, aminopenicillins, third-generation cephalosporins, and methicillin was classified as (and compared to) genotypic β-lactam resistance. Phenotypic macrolide resistance was compared with the relative abundance of ARGs conferring resistance to MLSB.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.