A workflow for the detection of antibiotic residues, measurement of water chemistry and preservation of hospital sink drain samples for metagenomic sequencing

Abstract

Background: Hospital sinks are environmental reservoirs that harbour healthcare-associated (HCA) pathogens. Selective pressures in sink environments, such as antibiotic residues, nutrient waste and hardness ions, may promote antibiotic resistance gene (ARG) exchange between bacteria. However, cheap and accurate sampling methods to characterize these factors are lacking.

Aims: To validate a workflow to detect antibiotic residues and evaluate water chemistry using dipsticks. Secondarily, to validate boric acid to preserve the taxonomic and ARG ('resistome') composition of sink trap samples for metagenomic sequencing.

Methods: Antibiotic residue dipsticks were validated against serial dilutions of ampicillin, doxycycline, sulfamethoxazole and ciprofloxacin, and water chemistry dipsticks against serial dilutions of chemical calibration standards. Sink trap aspirates were used for a 'real-world' pilot evaluation of dipsticks. To assess boric acid as a preservative of microbial diversity, the impact of incubation with and without boric acid at ∼22 °C on metagenomic sequencing outputs was evaluated at Day 2 and Day 5 compared with baseline (Day 0).

Findings: The limits of detection for each antibiotic were: 3 μg/L (ampicillin), 10 μg/L (doxycycline), 20 μg/L (sulfamethoxazole) and 8 μg/L (ciprofloxacin). The best performing water chemistry dipstick correctly characterized 34/40 (85%) standards in a concentration-dependent manner. One trap sample tested positive for the presence of tetracyclines and sulphonamides. Taxonomic and resistome composition were largely maintained after storage with boric acid at ∼22 °C for up to five days.

Conclusions: Dipsticks can be used to detect antibiotic residues and characterize water chemistry in sink trap samples. Boric acid was an effective preservative of trap sample composition, representing a low-cost alternative to cold-chain transport.

Keywords: Antibiotic residues; Antimicrobial resistance; Hospital sinks; Water chemistry.

Figure 1

Antibiotic residues of tetracycline and sulphonamide classes of antibiotics detected in sink trap sample C. A control line indicates the dipstick is performing; antibiotics are characterized as present if the antibiotic-specific line is less dark than the control line or absent.

Figure 2

Taxonomic distributions in samples with and without boric acid. Top panel: relative abundances of phyla across sink samples faceted by sink (A, B, C), processing day and use of boric acid as a preservative. ≤ represents phyla under 0.01 proportion of reads. Bottom panel: mean absolute percentage error (MAPE) of taxonomic abundances between baseline (Day 0) and processing day faceted by sink and stratified by whether boric acid was used as a preservative, where 1-MAPE = 1 indicates perfect preservation of baseline taxonomic distributions.

Figure 3

Antimicrobial resistance gene (ARG) presence in samples with and without boric acid. Top panel: ARG reads identified in sink samples A–C by day of evaluation and with/without boric acid, represented as the proportion of reads conferring resistance to a given antimicrobial class, with the number of reads identified in each facet. ≤ represents resistances under 0.05 proportion of reads. Bottom panel: mean absolute percentage error (MAPE) of abundances between baseline (Day 0) and processing day faceted by sink sample A–C and stratified by whether boric acid was used as a preservative, where 1-MAPE = 1 indicates perfect preservation of baseline ARG proportions.